{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9997280391623606,
  "eval_steps": 500,
  "global_step": 7352,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00013598041881968996,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0,
      "loss": 0.973,
      "memory/device_mem_reserved(gib)": 47.64,
      "memory/max_mem_active(gib)": 37.56,
      "memory/max_mem_allocated(gib)": 37.56,
      "step": 1
    },
    {
      "epoch": 0.0002719608376393799,
      "grad_norm": 1.9296875,
      "learning_rate": 2.7210884353741497e-08,
      "loss": 1.109,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2
    },
    {
      "epoch": 0.0004079412564590699,
      "grad_norm": 0.9765625,
      "learning_rate": 5.4421768707482993e-08,
      "loss": 0.582,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3
    },
    {
      "epoch": 0.0005439216752787598,
      "grad_norm": 1.625,
      "learning_rate": 8.163265306122451e-08,
      "loss": 1.0596,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4
    },
    {
      "epoch": 0.0006799020940984499,
      "grad_norm": 0.93359375,
      "learning_rate": 1.0884353741496599e-07,
      "loss": 0.5951,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5
    },
    {
      "epoch": 0.0008158825129181398,
      "grad_norm": 1.4453125,
      "learning_rate": 1.360544217687075e-07,
      "loss": 0.9098,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6
    },
    {
      "epoch": 0.0009518629317378297,
      "grad_norm": 1.6796875,
      "learning_rate": 1.6326530612244901e-07,
      "loss": 1.0146,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7
    },
    {
      "epoch": 0.0010878433505575197,
      "grad_norm": 1.1875,
      "learning_rate": 1.904761904761905e-07,
      "loss": 0.6969,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 8
    },
    {
      "epoch": 0.0012238237693772097,
      "grad_norm": 1.1953125,
      "learning_rate": 2.1768707482993197e-07,
      "loss": 0.8058,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 9
    },
    {
      "epoch": 0.0013598041881968997,
      "grad_norm": 1.2734375,
      "learning_rate": 2.4489795918367347e-07,
      "loss": 0.8242,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 10
    },
    {
      "epoch": 0.0014957846070165896,
      "grad_norm": 0.84375,
      "learning_rate": 2.72108843537415e-07,
      "loss": 0.499,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 11
    },
    {
      "epoch": 0.0016317650258362796,
      "grad_norm": 1.578125,
      "learning_rate": 2.9931972789115645e-07,
      "loss": 0.9258,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 12
    },
    {
      "epoch": 0.0017677454446559696,
      "grad_norm": 1.1796875,
      "learning_rate": 3.2653061224489803e-07,
      "loss": 0.8031,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 13
    },
    {
      "epoch": 0.0019037258634756595,
      "grad_norm": 2.265625,
      "learning_rate": 3.537414965986395e-07,
      "loss": 1.1639,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 14
    },
    {
      "epoch": 0.0020397062822953495,
      "grad_norm": 1.5546875,
      "learning_rate": 3.80952380952381e-07,
      "loss": 0.959,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 15
    },
    {
      "epoch": 0.0021756867011150393,
      "grad_norm": 1.078125,
      "learning_rate": 4.0816326530612243e-07,
      "loss": 0.7271,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 16
    },
    {
      "epoch": 0.0023116671199347296,
      "grad_norm": 1.2265625,
      "learning_rate": 4.3537414965986395e-07,
      "loss": 0.8768,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 17
    },
    {
      "epoch": 0.0024476475387544194,
      "grad_norm": 1.5390625,
      "learning_rate": 4.6258503401360547e-07,
      "loss": 1.0246,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 18
    },
    {
      "epoch": 0.002583627957574109,
      "grad_norm": 1.21875,
      "learning_rate": 4.897959183673469e-07,
      "loss": 0.7928,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 19
    },
    {
      "epoch": 0.0027196083763937995,
      "grad_norm": 8.4375,
      "learning_rate": 5.170068027210885e-07,
      "loss": 1.9658,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 20
    },
    {
      "epoch": 0.0028555887952134893,
      "grad_norm": 1.203125,
      "learning_rate": 5.4421768707483e-07,
      "loss": 0.7484,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 21
    },
    {
      "epoch": 0.002991569214033179,
      "grad_norm": 1.078125,
      "learning_rate": 5.714285714285715e-07,
      "loss": 0.6932,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 22
    },
    {
      "epoch": 0.0031275496328528694,
      "grad_norm": 0.875,
      "learning_rate": 5.986394557823129e-07,
      "loss": 0.5314,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 23
    },
    {
      "epoch": 0.003263530051672559,
      "grad_norm": 1.2890625,
      "learning_rate": 6.258503401360545e-07,
      "loss": 0.8307,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 24
    },
    {
      "epoch": 0.003399510470492249,
      "grad_norm": 0.8203125,
      "learning_rate": 6.530612244897961e-07,
      "loss": 0.5645,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 25
    },
    {
      "epoch": 0.0035354908893119393,
      "grad_norm": 1.75,
      "learning_rate": 6.802721088435376e-07,
      "loss": 1.1427,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 26
    },
    {
      "epoch": 0.003671471308131629,
      "grad_norm": 1.546875,
      "learning_rate": 7.07482993197279e-07,
      "loss": 0.9694,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 27
    },
    {
      "epoch": 0.003807451726951319,
      "grad_norm": 1.328125,
      "learning_rate": 7.346938775510205e-07,
      "loss": 0.8584,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 28
    },
    {
      "epoch": 0.003943432145771009,
      "grad_norm": 1.5,
      "learning_rate": 7.61904761904762e-07,
      "loss": 0.9222,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 29
    },
    {
      "epoch": 0.004079412564590699,
      "grad_norm": 1.4296875,
      "learning_rate": 7.891156462585034e-07,
      "loss": 0.8164,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 30
    },
    {
      "epoch": 0.004215392983410389,
      "grad_norm": 1.7265625,
      "learning_rate": 8.163265306122449e-07,
      "loss": 0.9699,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 31
    },
    {
      "epoch": 0.004351373402230079,
      "grad_norm": 1.21875,
      "learning_rate": 8.435374149659864e-07,
      "loss": 0.7445,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 32
    },
    {
      "epoch": 0.0044873538210497685,
      "grad_norm": 1.390625,
      "learning_rate": 8.707482993197279e-07,
      "loss": 0.925,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 33
    },
    {
      "epoch": 0.004623334239869459,
      "grad_norm": 1.5390625,
      "learning_rate": 8.979591836734694e-07,
      "loss": 0.9312,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 34
    },
    {
      "epoch": 0.004759314658689149,
      "grad_norm": 1.953125,
      "learning_rate": 9.251700680272109e-07,
      "loss": 1.0957,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 35
    },
    {
      "epoch": 0.004895295077508839,
      "grad_norm": 1.3359375,
      "learning_rate": 9.523809523809525e-07,
      "loss": 0.9211,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 36
    },
    {
      "epoch": 0.005031275496328529,
      "grad_norm": 1.515625,
      "learning_rate": 9.795918367346939e-07,
      "loss": 0.9463,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 37
    },
    {
      "epoch": 0.005167255915148218,
      "grad_norm": 1.3203125,
      "learning_rate": 1.0068027210884354e-06,
      "loss": 0.8509,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 38
    },
    {
      "epoch": 0.005303236333967908,
      "grad_norm": 1.09375,
      "learning_rate": 1.034013605442177e-06,
      "loss": 0.7588,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 39
    },
    {
      "epoch": 0.005439216752787599,
      "grad_norm": 1.5390625,
      "learning_rate": 1.0612244897959184e-06,
      "loss": 0.9951,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 40
    },
    {
      "epoch": 0.005575197171607289,
      "grad_norm": 1.625,
      "learning_rate": 1.08843537414966e-06,
      "loss": 1.0251,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 41
    },
    {
      "epoch": 0.005711177590426979,
      "grad_norm": 1.15625,
      "learning_rate": 1.1156462585034015e-06,
      "loss": 0.7757,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 42
    },
    {
      "epoch": 0.005847158009246668,
      "grad_norm": 1.265625,
      "learning_rate": 1.142857142857143e-06,
      "loss": 0.7581,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 43
    },
    {
      "epoch": 0.005983138428066358,
      "grad_norm": 1.3046875,
      "learning_rate": 1.1700680272108845e-06,
      "loss": 0.8813,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 44
    },
    {
      "epoch": 0.006119118846886048,
      "grad_norm": 1.484375,
      "learning_rate": 1.1972789115646258e-06,
      "loss": 0.9515,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 45
    },
    {
      "epoch": 0.006255099265705739,
      "grad_norm": 1.1796875,
      "learning_rate": 1.2244897959183673e-06,
      "loss": 0.8459,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 46
    },
    {
      "epoch": 0.0063910796845254286,
      "grad_norm": 0.890625,
      "learning_rate": 1.251700680272109e-06,
      "loss": 0.6001,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 47
    },
    {
      "epoch": 0.006527060103345118,
      "grad_norm": 1.09375,
      "learning_rate": 1.2789115646258504e-06,
      "loss": 0.7272,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 48
    },
    {
      "epoch": 0.006663040522164808,
      "grad_norm": 1.578125,
      "learning_rate": 1.3061224489795921e-06,
      "loss": 1.0622,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 49
    },
    {
      "epoch": 0.006799020940984498,
      "grad_norm": 1.0703125,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 0.6992,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 50
    },
    {
      "epoch": 0.006935001359804188,
      "grad_norm": 1.265625,
      "learning_rate": 1.3605442176870751e-06,
      "loss": 0.7513,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 51
    },
    {
      "epoch": 0.0070709817786238785,
      "grad_norm": 1.1953125,
      "learning_rate": 1.3877551020408165e-06,
      "loss": 0.8815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 52
    },
    {
      "epoch": 0.007206962197443568,
      "grad_norm": 1.15625,
      "learning_rate": 1.414965986394558e-06,
      "loss": 0.7988,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 53
    },
    {
      "epoch": 0.007342942616263258,
      "grad_norm": 0.7890625,
      "learning_rate": 1.4421768707482995e-06,
      "loss": 0.5962,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 54
    },
    {
      "epoch": 0.007478923035082948,
      "grad_norm": 1.46875,
      "learning_rate": 1.469387755102041e-06,
      "loss": 1.0454,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 55
    },
    {
      "epoch": 0.007614903453902638,
      "grad_norm": 1.5390625,
      "learning_rate": 1.4965986394557825e-06,
      "loss": 1.0671,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 56
    },
    {
      "epoch": 0.007750883872722328,
      "grad_norm": 0.71484375,
      "learning_rate": 1.523809523809524e-06,
      "loss": 0.5267,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 57
    },
    {
      "epoch": 0.007886864291542018,
      "grad_norm": 1.0546875,
      "learning_rate": 1.5510204081632654e-06,
      "loss": 0.8149,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 58
    },
    {
      "epoch": 0.008022844710361708,
      "grad_norm": 1.515625,
      "learning_rate": 1.5782312925170069e-06,
      "loss": 1.0713,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 59
    },
    {
      "epoch": 0.008158825129181398,
      "grad_norm": 1.2734375,
      "learning_rate": 1.6054421768707484e-06,
      "loss": 0.9425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 60
    },
    {
      "epoch": 0.008294805548001088,
      "grad_norm": 1.40625,
      "learning_rate": 1.6326530612244897e-06,
      "loss": 1.0483,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 61
    },
    {
      "epoch": 0.008430785966820778,
      "grad_norm": 1.078125,
      "learning_rate": 1.6598639455782314e-06,
      "loss": 0.8156,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 62
    },
    {
      "epoch": 0.008566766385640467,
      "grad_norm": 1.328125,
      "learning_rate": 1.6870748299319727e-06,
      "loss": 0.8403,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 63
    },
    {
      "epoch": 0.008702746804460157,
      "grad_norm": 1.3984375,
      "learning_rate": 1.7142857142857145e-06,
      "loss": 1.0612,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 64
    },
    {
      "epoch": 0.008838727223279847,
      "grad_norm": 1.2109375,
      "learning_rate": 1.7414965986394558e-06,
      "loss": 0.8657,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 65
    },
    {
      "epoch": 0.008974707642099537,
      "grad_norm": 1.078125,
      "learning_rate": 1.7687074829931975e-06,
      "loss": 0.7617,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 66
    },
    {
      "epoch": 0.009110688060919228,
      "grad_norm": 1.4765625,
      "learning_rate": 1.7959183673469388e-06,
      "loss": 1.0361,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 67
    },
    {
      "epoch": 0.009246668479738918,
      "grad_norm": 0.98046875,
      "learning_rate": 1.8231292517006803e-06,
      "loss": 0.7256,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 68
    },
    {
      "epoch": 0.009382648898558608,
      "grad_norm": 0.86328125,
      "learning_rate": 1.8503401360544219e-06,
      "loss": 0.62,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 69
    },
    {
      "epoch": 0.009518629317378298,
      "grad_norm": 0.96484375,
      "learning_rate": 1.8775510204081634e-06,
      "loss": 0.6696,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 70
    },
    {
      "epoch": 0.009654609736197988,
      "grad_norm": 1.25,
      "learning_rate": 1.904761904761905e-06,
      "loss": 0.8675,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 71
    },
    {
      "epoch": 0.009790590155017678,
      "grad_norm": 4.0,
      "learning_rate": 1.9319727891156464e-06,
      "loss": 1.1577,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 72
    },
    {
      "epoch": 0.009926570573837367,
      "grad_norm": 1.0625,
      "learning_rate": 1.9591836734693877e-06,
      "loss": 0.7135,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 73
    },
    {
      "epoch": 0.010062550992657057,
      "grad_norm": 1.0078125,
      "learning_rate": 1.9863945578231295e-06,
      "loss": 0.7959,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 74
    },
    {
      "epoch": 0.010198531411476747,
      "grad_norm": 0.88671875,
      "learning_rate": 2.0136054421768708e-06,
      "loss": 0.6812,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 75
    },
    {
      "epoch": 0.010334511830296437,
      "grad_norm": 1.1953125,
      "learning_rate": 2.0408163265306125e-06,
      "loss": 0.9491,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 76
    },
    {
      "epoch": 0.010470492249116127,
      "grad_norm": 1.0859375,
      "learning_rate": 2.068027210884354e-06,
      "loss": 0.8219,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 77
    },
    {
      "epoch": 0.010606472667935817,
      "grad_norm": 1.0859375,
      "learning_rate": 2.0952380952380955e-06,
      "loss": 0.887,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 78
    },
    {
      "epoch": 0.010742453086755508,
      "grad_norm": 1.1640625,
      "learning_rate": 2.122448979591837e-06,
      "loss": 0.9248,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 79
    },
    {
      "epoch": 0.010878433505575198,
      "grad_norm": 1.4296875,
      "learning_rate": 2.1496598639455786e-06,
      "loss": 1.009,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 80
    },
    {
      "epoch": 0.011014413924394888,
      "grad_norm": 0.94140625,
      "learning_rate": 2.17687074829932e-06,
      "loss": 0.7104,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 81
    },
    {
      "epoch": 0.011150394343214578,
      "grad_norm": 1.15625,
      "learning_rate": 2.2040816326530616e-06,
      "loss": 0.9108,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 82
    },
    {
      "epoch": 0.011286374762034267,
      "grad_norm": 0.79296875,
      "learning_rate": 2.231292517006803e-06,
      "loss": 0.6615,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 83
    },
    {
      "epoch": 0.011422355180853957,
      "grad_norm": 0.96875,
      "learning_rate": 2.2585034013605447e-06,
      "loss": 0.7593,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 84
    },
    {
      "epoch": 0.011558335599673647,
      "grad_norm": 1.15625,
      "learning_rate": 2.285714285714286e-06,
      "loss": 1.0078,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 85
    },
    {
      "epoch": 0.011694316018493337,
      "grad_norm": 0.94921875,
      "learning_rate": 2.3129251700680273e-06,
      "loss": 0.7004,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 86
    },
    {
      "epoch": 0.011830296437313027,
      "grad_norm": 0.5078125,
      "learning_rate": 2.340136054421769e-06,
      "loss": 0.4699,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 87
    },
    {
      "epoch": 0.011966276856132716,
      "grad_norm": 0.828125,
      "learning_rate": 2.3673469387755103e-06,
      "loss": 0.8008,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 88
    },
    {
      "epoch": 0.012102257274952406,
      "grad_norm": 1.0703125,
      "learning_rate": 2.3945578231292516e-06,
      "loss": 0.8563,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 89
    },
    {
      "epoch": 0.012238237693772096,
      "grad_norm": 0.72265625,
      "learning_rate": 2.4217687074829934e-06,
      "loss": 0.6979,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 90
    },
    {
      "epoch": 0.012374218112591786,
      "grad_norm": 2.09375,
      "learning_rate": 2.4489795918367347e-06,
      "loss": 0.7786,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 91
    },
    {
      "epoch": 0.012510198531411477,
      "grad_norm": 0.6328125,
      "learning_rate": 2.4761904761904764e-06,
      "loss": 0.6012,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 92
    },
    {
      "epoch": 0.012646178950231167,
      "grad_norm": 0.72265625,
      "learning_rate": 2.503401360544218e-06,
      "loss": 0.6615,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 93
    },
    {
      "epoch": 0.012782159369050857,
      "grad_norm": 0.96484375,
      "learning_rate": 2.530612244897959e-06,
      "loss": 0.9347,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 94
    },
    {
      "epoch": 0.012918139787870547,
      "grad_norm": 0.95703125,
      "learning_rate": 2.5578231292517007e-06,
      "loss": 0.8934,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 95
    },
    {
      "epoch": 0.013054120206690237,
      "grad_norm": 1.0625,
      "learning_rate": 2.5850340136054425e-06,
      "loss": 0.7243,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 96
    },
    {
      "epoch": 0.013190100625509927,
      "grad_norm": 0.9375,
      "learning_rate": 2.6122448979591842e-06,
      "loss": 0.9403,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 97
    },
    {
      "epoch": 0.013326081044329616,
      "grad_norm": 0.6953125,
      "learning_rate": 2.639455782312925e-06,
      "loss": 0.6966,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 98
    },
    {
      "epoch": 0.013462061463149306,
      "grad_norm": 0.640625,
      "learning_rate": 2.666666666666667e-06,
      "loss": 0.7039,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 99
    },
    {
      "epoch": 0.013598041881968996,
      "grad_norm": 0.69140625,
      "learning_rate": 2.6938775510204086e-06,
      "loss": 0.6383,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 100
    },
    {
      "epoch": 0.013734022300788686,
      "grad_norm": 0.94140625,
      "learning_rate": 2.7210884353741503e-06,
      "loss": 0.8105,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 101
    },
    {
      "epoch": 0.013870002719608376,
      "grad_norm": 0.78125,
      "learning_rate": 2.748299319727891e-06,
      "loss": 0.6768,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 102
    },
    {
      "epoch": 0.014005983138428066,
      "grad_norm": 0.75,
      "learning_rate": 2.775510204081633e-06,
      "loss": 0.7967,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 103
    },
    {
      "epoch": 0.014141963557247757,
      "grad_norm": 0.57421875,
      "learning_rate": 2.8027210884353746e-06,
      "loss": 0.6797,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 104
    },
    {
      "epoch": 0.014277943976067447,
      "grad_norm": 0.94140625,
      "learning_rate": 2.829931972789116e-06,
      "loss": 1.04,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 105
    },
    {
      "epoch": 0.014413924394887137,
      "grad_norm": 0.80078125,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 0.8249,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 106
    },
    {
      "epoch": 0.014549904813706827,
      "grad_norm": 0.7421875,
      "learning_rate": 2.884353741496599e-06,
      "loss": 0.9142,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 107
    },
    {
      "epoch": 0.014685885232526516,
      "grad_norm": 0.7421875,
      "learning_rate": 2.9115646258503403e-06,
      "loss": 0.9105,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 108
    },
    {
      "epoch": 0.014821865651346206,
      "grad_norm": 2.375,
      "learning_rate": 2.938775510204082e-06,
      "loss": 0.8752,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 109
    },
    {
      "epoch": 0.014957846070165896,
      "grad_norm": 0.734375,
      "learning_rate": 2.9659863945578233e-06,
      "loss": 0.7622,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 110
    },
    {
      "epoch": 0.015093826488985586,
      "grad_norm": 0.66015625,
      "learning_rate": 2.993197278911565e-06,
      "loss": 0.6536,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 111
    },
    {
      "epoch": 0.015229806907805276,
      "grad_norm": 0.66015625,
      "learning_rate": 3.0204081632653064e-06,
      "loss": 0.7599,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 112
    },
    {
      "epoch": 0.015365787326624965,
      "grad_norm": 0.81640625,
      "learning_rate": 3.047619047619048e-06,
      "loss": 0.7565,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 113
    },
    {
      "epoch": 0.015501767745444655,
      "grad_norm": 0.56640625,
      "learning_rate": 3.0748299319727894e-06,
      "loss": 0.7336,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 114
    },
    {
      "epoch": 0.015637748164264345,
      "grad_norm": 0.86328125,
      "learning_rate": 3.1020408163265307e-06,
      "loss": 0.9966,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 115
    },
    {
      "epoch": 0.015773728583084037,
      "grad_norm": 0.60546875,
      "learning_rate": 3.1292517006802725e-06,
      "loss": 0.9222,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 116
    },
    {
      "epoch": 0.015909709001903725,
      "grad_norm": 0.53515625,
      "learning_rate": 3.1564625850340138e-06,
      "loss": 0.7316,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 117
    },
    {
      "epoch": 0.016045689420723416,
      "grad_norm": 1.78125,
      "learning_rate": 3.183673469387755e-06,
      "loss": 0.7899,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 118
    },
    {
      "epoch": 0.016181669839543104,
      "grad_norm": 0.453125,
      "learning_rate": 3.210884353741497e-06,
      "loss": 0.5801,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 119
    },
    {
      "epoch": 0.016317650258362796,
      "grad_norm": 0.71875,
      "learning_rate": 3.2380952380952385e-06,
      "loss": 0.9632,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 120
    },
    {
      "epoch": 0.016453630677182484,
      "grad_norm": 0.73046875,
      "learning_rate": 3.2653061224489794e-06,
      "loss": 0.5618,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 121
    },
    {
      "epoch": 0.016589611096002176,
      "grad_norm": 0.89453125,
      "learning_rate": 3.292517006802721e-06,
      "loss": 1.0436,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 122
    },
    {
      "epoch": 0.016725591514821867,
      "grad_norm": 0.7734375,
      "learning_rate": 3.319727891156463e-06,
      "loss": 0.9748,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 123
    },
    {
      "epoch": 0.016861571933641555,
      "grad_norm": 0.2431640625,
      "learning_rate": 3.3469387755102046e-06,
      "loss": 0.3447,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 124
    },
    {
      "epoch": 0.016997552352461247,
      "grad_norm": 0.62109375,
      "learning_rate": 3.3741496598639455e-06,
      "loss": 0.8794,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 125
    },
    {
      "epoch": 0.017133532771280935,
      "grad_norm": 0.5390625,
      "learning_rate": 3.4013605442176872e-06,
      "loss": 0.7708,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 126
    },
    {
      "epoch": 0.017269513190100626,
      "grad_norm": 0.6875,
      "learning_rate": 3.428571428571429e-06,
      "loss": 0.9043,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 127
    },
    {
      "epoch": 0.017405493608920315,
      "grad_norm": 0.3671875,
      "learning_rate": 3.4557823129251707e-06,
      "loss": 0.5068,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 128
    },
    {
      "epoch": 0.017541474027740006,
      "grad_norm": 0.59765625,
      "learning_rate": 3.4829931972789116e-06,
      "loss": 0.7539,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 129
    },
    {
      "epoch": 0.017677454446559694,
      "grad_norm": 0.546875,
      "learning_rate": 3.5102040816326533e-06,
      "loss": 0.923,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 130
    },
    {
      "epoch": 0.017813434865379386,
      "grad_norm": 0.46484375,
      "learning_rate": 3.537414965986395e-06,
      "loss": 0.6766,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 131
    },
    {
      "epoch": 0.017949415284199074,
      "grad_norm": 0.7265625,
      "learning_rate": 3.5646258503401363e-06,
      "loss": 0.8231,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 132
    },
    {
      "epoch": 0.018085395703018765,
      "grad_norm": 0.49609375,
      "learning_rate": 3.5918367346938777e-06,
      "loss": 0.7767,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 133
    },
    {
      "epoch": 0.018221376121838457,
      "grad_norm": 0.9296875,
      "learning_rate": 3.6190476190476194e-06,
      "loss": 0.7992,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 134
    },
    {
      "epoch": 0.018357356540658145,
      "grad_norm": 0.419921875,
      "learning_rate": 3.6462585034013607e-06,
      "loss": 0.6965,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 135
    },
    {
      "epoch": 0.018493336959477837,
      "grad_norm": 0.5703125,
      "learning_rate": 3.6734693877551024e-06,
      "loss": 0.6017,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 136
    },
    {
      "epoch": 0.018629317378297525,
      "grad_norm": 0.70703125,
      "learning_rate": 3.7006802721088437e-06,
      "loss": 0.6126,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 137
    },
    {
      "epoch": 0.018765297797117216,
      "grad_norm": 0.4765625,
      "learning_rate": 3.7278911564625855e-06,
      "loss": 0.6598,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 138
    },
    {
      "epoch": 0.018901278215936904,
      "grad_norm": 0.75390625,
      "learning_rate": 3.7551020408163268e-06,
      "loss": 0.8512,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 139
    },
    {
      "epoch": 0.019037258634756596,
      "grad_norm": 0.7109375,
      "learning_rate": 3.7823129251700685e-06,
      "loss": 0.7428,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 140
    },
    {
      "epoch": 0.019173239053576284,
      "grad_norm": 0.419921875,
      "learning_rate": 3.80952380952381e-06,
      "loss": 0.7266,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 141
    },
    {
      "epoch": 0.019309219472395975,
      "grad_norm": 0.51953125,
      "learning_rate": 3.836734693877551e-06,
      "loss": 0.6245,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 142
    },
    {
      "epoch": 0.019445199891215664,
      "grad_norm": 0.63671875,
      "learning_rate": 3.863945578231293e-06,
      "loss": 0.8674,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 143
    },
    {
      "epoch": 0.019581180310035355,
      "grad_norm": 0.390625,
      "learning_rate": 3.891156462585035e-06,
      "loss": 0.6572,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 144
    },
    {
      "epoch": 0.019717160728855043,
      "grad_norm": 0.439453125,
      "learning_rate": 3.9183673469387755e-06,
      "loss": 0.8815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 145
    },
    {
      "epoch": 0.019853141147674735,
      "grad_norm": 0.546875,
      "learning_rate": 3.945578231292517e-06,
      "loss": 0.8299,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 146
    },
    {
      "epoch": 0.019989121566494426,
      "grad_norm": 0.5703125,
      "learning_rate": 3.972789115646259e-06,
      "loss": 0.9198,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 147
    },
    {
      "epoch": 0.020125101985314114,
      "grad_norm": 0.53125,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.8075,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 148
    },
    {
      "epoch": 0.020261082404133806,
      "grad_norm": 1.0234375,
      "learning_rate": 4.0272108843537416e-06,
      "loss": 0.7716,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 149
    },
    {
      "epoch": 0.020397062822953494,
      "grad_norm": 0.52734375,
      "learning_rate": 4.054421768707483e-06,
      "loss": 0.862,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 150
    },
    {
      "epoch": 0.020533043241773186,
      "grad_norm": 0.78515625,
      "learning_rate": 4.081632653061225e-06,
      "loss": 0.7142,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 151
    },
    {
      "epoch": 0.020669023660592874,
      "grad_norm": 0.412109375,
      "learning_rate": 4.108843537414967e-06,
      "loss": 0.6968,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 152
    },
    {
      "epoch": 0.020805004079412565,
      "grad_norm": 0.52734375,
      "learning_rate": 4.136054421768708e-06,
      "loss": 0.8211,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 153
    },
    {
      "epoch": 0.020940984498232253,
      "grad_norm": 0.380859375,
      "learning_rate": 4.163265306122449e-06,
      "loss": 0.7043,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 154
    },
    {
      "epoch": 0.021076964917051945,
      "grad_norm": 0.4375,
      "learning_rate": 4.190476190476191e-06,
      "loss": 0.6123,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 155
    },
    {
      "epoch": 0.021212945335871633,
      "grad_norm": 0.427734375,
      "learning_rate": 4.217687074829933e-06,
      "loss": 0.757,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 156
    },
    {
      "epoch": 0.021348925754691325,
      "grad_norm": 0.42578125,
      "learning_rate": 4.244897959183674e-06,
      "loss": 0.7404,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 157
    },
    {
      "epoch": 0.021484906173511016,
      "grad_norm": 0.3203125,
      "learning_rate": 4.2721088435374154e-06,
      "loss": 0.6502,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 158
    },
    {
      "epoch": 0.021620886592330704,
      "grad_norm": 0.39453125,
      "learning_rate": 4.299319727891157e-06,
      "loss": 0.7308,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 159
    },
    {
      "epoch": 0.021756867011150396,
      "grad_norm": 0.39453125,
      "learning_rate": 4.326530612244899e-06,
      "loss": 0.7619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 160
    },
    {
      "epoch": 0.021892847429970084,
      "grad_norm": 0.48046875,
      "learning_rate": 4.35374149659864e-06,
      "loss": 0.9688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 161
    },
    {
      "epoch": 0.022028827848789775,
      "grad_norm": 0.404296875,
      "learning_rate": 4.3809523809523815e-06,
      "loss": 0.6945,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 162
    },
    {
      "epoch": 0.022164808267609463,
      "grad_norm": 0.3359375,
      "learning_rate": 4.408163265306123e-06,
      "loss": 0.6489,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 163
    },
    {
      "epoch": 0.022300788686429155,
      "grad_norm": 0.7890625,
      "learning_rate": 4.435374149659865e-06,
      "loss": 0.8934,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 164
    },
    {
      "epoch": 0.022436769105248843,
      "grad_norm": 1.25,
      "learning_rate": 4.462585034013606e-06,
      "loss": 0.7466,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 165
    },
    {
      "epoch": 0.022572749524068535,
      "grad_norm": 0.373046875,
      "learning_rate": 4.489795918367348e-06,
      "loss": 0.8389,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 166
    },
    {
      "epoch": 0.022708729942888223,
      "grad_norm": 0.47265625,
      "learning_rate": 4.517006802721089e-06,
      "loss": 0.6725,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 167
    },
    {
      "epoch": 0.022844710361707914,
      "grad_norm": 0.43359375,
      "learning_rate": 4.54421768707483e-06,
      "loss": 0.7062,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 168
    },
    {
      "epoch": 0.022980690780527602,
      "grad_norm": 0.51953125,
      "learning_rate": 4.571428571428572e-06,
      "loss": 0.5658,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 169
    },
    {
      "epoch": 0.023116671199347294,
      "grad_norm": 0.447265625,
      "learning_rate": 4.598639455782314e-06,
      "loss": 0.8704,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 170
    },
    {
      "epoch": 0.023252651618166986,
      "grad_norm": 0.396484375,
      "learning_rate": 4.6258503401360546e-06,
      "loss": 0.6842,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 171
    },
    {
      "epoch": 0.023388632036986674,
      "grad_norm": 0.478515625,
      "learning_rate": 4.653061224489796e-06,
      "loss": 0.7699,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 172
    },
    {
      "epoch": 0.023524612455806365,
      "grad_norm": 0.4375,
      "learning_rate": 4.680272108843538e-06,
      "loss": 0.6392,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 173
    },
    {
      "epoch": 0.023660592874626053,
      "grad_norm": 0.53515625,
      "learning_rate": 4.707482993197279e-06,
      "loss": 0.7949,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 174
    },
    {
      "epoch": 0.023796573293445745,
      "grad_norm": 0.28125,
      "learning_rate": 4.734693877551021e-06,
      "loss": 0.5171,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 175
    },
    {
      "epoch": 0.023932553712265433,
      "grad_norm": 0.6328125,
      "learning_rate": 4.761904761904762e-06,
      "loss": 0.9404,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 176
    },
    {
      "epoch": 0.024068534131085124,
      "grad_norm": 0.3203125,
      "learning_rate": 4.789115646258503e-06,
      "loss": 0.6393,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 177
    },
    {
      "epoch": 0.024204514549904813,
      "grad_norm": 0.40234375,
      "learning_rate": 4.816326530612245e-06,
      "loss": 0.8898,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 178
    },
    {
      "epoch": 0.024340494968724504,
      "grad_norm": 0.30859375,
      "learning_rate": 4.843537414965987e-06,
      "loss": 0.4834,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 179
    },
    {
      "epoch": 0.024476475387544192,
      "grad_norm": 0.82421875,
      "learning_rate": 4.8707482993197285e-06,
      "loss": 0.7681,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 180
    },
    {
      "epoch": 0.024612455806363884,
      "grad_norm": 0.55078125,
      "learning_rate": 4.897959183673469e-06,
      "loss": 0.7754,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 181
    },
    {
      "epoch": 0.024748436225183572,
      "grad_norm": 0.341796875,
      "learning_rate": 4.925170068027211e-06,
      "loss": 0.6714,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 182
    },
    {
      "epoch": 0.024884416644003263,
      "grad_norm": 0.369140625,
      "learning_rate": 4.952380952380953e-06,
      "loss": 0.7147,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 183
    },
    {
      "epoch": 0.025020397062822955,
      "grad_norm": 0.43359375,
      "learning_rate": 4.979591836734694e-06,
      "loss": 0.6795,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 184
    },
    {
      "epoch": 0.025156377481642643,
      "grad_norm": 0.42578125,
      "learning_rate": 5.006802721088436e-06,
      "loss": 0.8494,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 185
    },
    {
      "epoch": 0.025292357900462335,
      "grad_norm": 0.408203125,
      "learning_rate": 5.034013605442177e-06,
      "loss": 0.5566,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 186
    },
    {
      "epoch": 0.025428338319282023,
      "grad_norm": 0.65234375,
      "learning_rate": 5.061224489795918e-06,
      "loss": 0.6413,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 187
    },
    {
      "epoch": 0.025564318738101714,
      "grad_norm": 0.439453125,
      "learning_rate": 5.088435374149661e-06,
      "loss": 0.8281,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 188
    },
    {
      "epoch": 0.025700299156921402,
      "grad_norm": 0.400390625,
      "learning_rate": 5.1156462585034015e-06,
      "loss": 0.8175,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 189
    },
    {
      "epoch": 0.025836279575741094,
      "grad_norm": 0.51953125,
      "learning_rate": 5.142857142857142e-06,
      "loss": 0.8748,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 190
    },
    {
      "epoch": 0.025972259994560782,
      "grad_norm": 0.703125,
      "learning_rate": 5.170068027210885e-06,
      "loss": 0.6808,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 191
    },
    {
      "epoch": 0.026108240413380474,
      "grad_norm": 0.21484375,
      "learning_rate": 5.197278911564626e-06,
      "loss": 0.3787,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 192
    },
    {
      "epoch": 0.02624422083220016,
      "grad_norm": 0.53125,
      "learning_rate": 5.2244897959183684e-06,
      "loss": 0.637,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 193
    },
    {
      "epoch": 0.026380201251019853,
      "grad_norm": 0.369140625,
      "learning_rate": 5.251700680272109e-06,
      "loss": 0.7282,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 194
    },
    {
      "epoch": 0.026516181669839545,
      "grad_norm": 0.484375,
      "learning_rate": 5.27891156462585e-06,
      "loss": 0.9263,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 195
    },
    {
      "epoch": 0.026652162088659233,
      "grad_norm": 0.455078125,
      "learning_rate": 5.306122448979593e-06,
      "loss": 0.8765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 196
    },
    {
      "epoch": 0.026788142507478924,
      "grad_norm": 0.341796875,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.6606,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 197
    },
    {
      "epoch": 0.026924122926298612,
      "grad_norm": 0.390625,
      "learning_rate": 5.3605442176870745e-06,
      "loss": 0.6154,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 198
    },
    {
      "epoch": 0.027060103345118304,
      "grad_norm": 0.66015625,
      "learning_rate": 5.387755102040817e-06,
      "loss": 0.8929,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 199
    },
    {
      "epoch": 0.027196083763937992,
      "grad_norm": 0.458984375,
      "learning_rate": 5.414965986394558e-06,
      "loss": 0.693,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 200
    },
    {
      "epoch": 0.027332064182757684,
      "grad_norm": 0.56640625,
      "learning_rate": 5.442176870748301e-06,
      "loss": 0.9438,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 201
    },
    {
      "epoch": 0.027468044601577372,
      "grad_norm": 0.57421875,
      "learning_rate": 5.4693877551020415e-06,
      "loss": 0.8498,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 202
    },
    {
      "epoch": 0.027604025020397063,
      "grad_norm": 0.408203125,
      "learning_rate": 5.496598639455782e-06,
      "loss": 0.6166,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 203
    },
    {
      "epoch": 0.02774000543921675,
      "grad_norm": 0.37890625,
      "learning_rate": 5.523809523809525e-06,
      "loss": 0.6255,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 204
    },
    {
      "epoch": 0.027875985858036443,
      "grad_norm": 0.59765625,
      "learning_rate": 5.551020408163266e-06,
      "loss": 0.9277,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 205
    },
    {
      "epoch": 0.02801196627685613,
      "grad_norm": 0.42578125,
      "learning_rate": 5.578231292517007e-06,
      "loss": 0.7293,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 206
    },
    {
      "epoch": 0.028147946695675823,
      "grad_norm": 0.3671875,
      "learning_rate": 5.605442176870749e-06,
      "loss": 0.6662,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 207
    },
    {
      "epoch": 0.028283927114495514,
      "grad_norm": 0.796875,
      "learning_rate": 5.63265306122449e-06,
      "loss": 0.8867,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 208
    },
    {
      "epoch": 0.028419907533315202,
      "grad_norm": 0.65234375,
      "learning_rate": 5.659863945578232e-06,
      "loss": 0.7655,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 209
    },
    {
      "epoch": 0.028555887952134894,
      "grad_norm": 0.52734375,
      "learning_rate": 5.687074829931974e-06,
      "loss": 0.8082,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 210
    },
    {
      "epoch": 0.028691868370954582,
      "grad_norm": 0.341796875,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 0.5872,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 211
    },
    {
      "epoch": 0.028827848789774273,
      "grad_norm": 0.376953125,
      "learning_rate": 5.741496598639456e-06,
      "loss": 0.7412,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 212
    },
    {
      "epoch": 0.02896382920859396,
      "grad_norm": 0.396484375,
      "learning_rate": 5.768707482993198e-06,
      "loss": 0.6835,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 213
    },
    {
      "epoch": 0.029099809627413653,
      "grad_norm": 0.47265625,
      "learning_rate": 5.795918367346939e-06,
      "loss": 0.5636,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 214
    },
    {
      "epoch": 0.02923579004623334,
      "grad_norm": 0.60546875,
      "learning_rate": 5.823129251700681e-06,
      "loss": 0.741,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 215
    },
    {
      "epoch": 0.029371770465053033,
      "grad_norm": 0.57421875,
      "learning_rate": 5.850340136054422e-06,
      "loss": 0.6878,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 216
    },
    {
      "epoch": 0.02950775088387272,
      "grad_norm": 0.56640625,
      "learning_rate": 5.877551020408164e-06,
      "loss": 0.7103,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 217
    },
    {
      "epoch": 0.029643731302692412,
      "grad_norm": 0.298828125,
      "learning_rate": 5.904761904761905e-06,
      "loss": 0.4969,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 218
    },
    {
      "epoch": 0.029779711721512104,
      "grad_norm": 0.67578125,
      "learning_rate": 5.931972789115647e-06,
      "loss": 0.6834,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 219
    },
    {
      "epoch": 0.029915692140331792,
      "grad_norm": 0.357421875,
      "learning_rate": 5.959183673469388e-06,
      "loss": 0.639,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 220
    },
    {
      "epoch": 0.030051672559151484,
      "grad_norm": 0.462890625,
      "learning_rate": 5.98639455782313e-06,
      "loss": 0.7853,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 221
    },
    {
      "epoch": 0.03018765297797117,
      "grad_norm": 0.453125,
      "learning_rate": 6.013605442176871e-06,
      "loss": 0.8154,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 222
    },
    {
      "epoch": 0.030323633396790863,
      "grad_norm": 0.443359375,
      "learning_rate": 6.040816326530613e-06,
      "loss": 0.734,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 223
    },
    {
      "epoch": 0.03045961381561055,
      "grad_norm": 0.515625,
      "learning_rate": 6.0680272108843545e-06,
      "loss": 0.5636,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 224
    },
    {
      "epoch": 0.030595594234430243,
      "grad_norm": 0.326171875,
      "learning_rate": 6.095238095238096e-06,
      "loss": 0.5664,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 225
    },
    {
      "epoch": 0.03073157465324993,
      "grad_norm": 0.44921875,
      "learning_rate": 6.122448979591837e-06,
      "loss": 0.8218,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 226
    },
    {
      "epoch": 0.030867555072069622,
      "grad_norm": 0.337890625,
      "learning_rate": 6.149659863945579e-06,
      "loss": 0.7049,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 227
    },
    {
      "epoch": 0.03100353549088931,
      "grad_norm": 0.33203125,
      "learning_rate": 6.1768707482993206e-06,
      "loss": 0.6589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 228
    },
    {
      "epoch": 0.031139515909709002,
      "grad_norm": 0.435546875,
      "learning_rate": 6.2040816326530614e-06,
      "loss": 0.694,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 229
    },
    {
      "epoch": 0.03127549632852869,
      "grad_norm": 0.51953125,
      "learning_rate": 6.231292517006803e-06,
      "loss": 0.5724,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 230
    },
    {
      "epoch": 0.031411476747348385,
      "grad_norm": 0.45703125,
      "learning_rate": 6.258503401360545e-06,
      "loss": 0.5073,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 231
    },
    {
      "epoch": 0.03154745716616807,
      "grad_norm": 0.341796875,
      "learning_rate": 6.285714285714286e-06,
      "loss": 0.6092,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 232
    },
    {
      "epoch": 0.03168343758498776,
      "grad_norm": 0.52734375,
      "learning_rate": 6.3129251700680275e-06,
      "loss": 0.7922,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 233
    },
    {
      "epoch": 0.03181941800380745,
      "grad_norm": 0.40625,
      "learning_rate": 6.340136054421769e-06,
      "loss": 0.889,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 234
    },
    {
      "epoch": 0.031955398422627145,
      "grad_norm": 0.45703125,
      "learning_rate": 6.36734693877551e-06,
      "loss": 0.7852,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 235
    },
    {
      "epoch": 0.03209137884144683,
      "grad_norm": 0.26171875,
      "learning_rate": 6.394557823129253e-06,
      "loss": 0.4699,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 236
    },
    {
      "epoch": 0.03222735926026652,
      "grad_norm": 0.7578125,
      "learning_rate": 6.421768707482994e-06,
      "loss": 0.61,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 237
    },
    {
      "epoch": 0.03236333967908621,
      "grad_norm": 0.478515625,
      "learning_rate": 6.4489795918367345e-06,
      "loss": 0.7663,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 238
    },
    {
      "epoch": 0.032499320097905904,
      "grad_norm": 0.341796875,
      "learning_rate": 6.476190476190477e-06,
      "loss": 0.508,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 239
    },
    {
      "epoch": 0.03263530051672559,
      "grad_norm": 0.59765625,
      "learning_rate": 6.503401360544218e-06,
      "loss": 0.6351,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 240
    },
    {
      "epoch": 0.03277128093554528,
      "grad_norm": 0.357421875,
      "learning_rate": 6.530612244897959e-06,
      "loss": 0.6646,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 241
    },
    {
      "epoch": 0.03290726135436497,
      "grad_norm": 0.5234375,
      "learning_rate": 6.557823129251701e-06,
      "loss": 0.9951,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 242
    },
    {
      "epoch": 0.03304324177318466,
      "grad_norm": 0.4140625,
      "learning_rate": 6.585034013605442e-06,
      "loss": 0.7405,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 243
    },
    {
      "epoch": 0.03317922219200435,
      "grad_norm": 0.439453125,
      "learning_rate": 6.612244897959185e-06,
      "loss": 0.883,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 244
    },
    {
      "epoch": 0.03331520261082404,
      "grad_norm": 0.40625,
      "learning_rate": 6.639455782312926e-06,
      "loss": 0.8263,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 245
    },
    {
      "epoch": 0.033451183029643734,
      "grad_norm": 0.56640625,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.6847,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 246
    },
    {
      "epoch": 0.03358716344846342,
      "grad_norm": 0.494140625,
      "learning_rate": 6.693877551020409e-06,
      "loss": 0.7989,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 247
    },
    {
      "epoch": 0.03372314386728311,
      "grad_norm": 0.3203125,
      "learning_rate": 6.72108843537415e-06,
      "loss": 0.4845,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 248
    },
    {
      "epoch": 0.0338591242861028,
      "grad_norm": 0.337890625,
      "learning_rate": 6.748299319727891e-06,
      "loss": 0.6258,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 249
    },
    {
      "epoch": 0.033995104704922494,
      "grad_norm": 0.330078125,
      "learning_rate": 6.7755102040816336e-06,
      "loss": 0.6061,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 250
    },
    {
      "epoch": 0.03413108512374218,
      "grad_norm": 0.4921875,
      "learning_rate": 6.8027210884353745e-06,
      "loss": 0.5195,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 251
    },
    {
      "epoch": 0.03426706554256187,
      "grad_norm": 0.32421875,
      "learning_rate": 6.829931972789117e-06,
      "loss": 0.6489,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 252
    },
    {
      "epoch": 0.03440304596138156,
      "grad_norm": 0.68359375,
      "learning_rate": 6.857142857142858e-06,
      "loss": 0.84,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 253
    },
    {
      "epoch": 0.03453902638020125,
      "grad_norm": 0.34375,
      "learning_rate": 6.884353741496599e-06,
      "loss": 0.6265,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 254
    },
    {
      "epoch": 0.03467500679902094,
      "grad_norm": 0.703125,
      "learning_rate": 6.911564625850341e-06,
      "loss": 0.9307,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 255
    },
    {
      "epoch": 0.03481098721784063,
      "grad_norm": 0.34765625,
      "learning_rate": 6.938775510204082e-06,
      "loss": 0.5264,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 256
    },
    {
      "epoch": 0.034946967636660324,
      "grad_norm": 0.4296875,
      "learning_rate": 6.965986394557823e-06,
      "loss": 0.7669,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 257
    },
    {
      "epoch": 0.03508294805548001,
      "grad_norm": 0.318359375,
      "learning_rate": 6.993197278911566e-06,
      "loss": 0.6453,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 258
    },
    {
      "epoch": 0.0352189284742997,
      "grad_norm": 0.279296875,
      "learning_rate": 7.020408163265307e-06,
      "loss": 0.4497,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 259
    },
    {
      "epoch": 0.03535490889311939,
      "grad_norm": 0.408203125,
      "learning_rate": 7.047619047619048e-06,
      "loss": 0.6089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 260
    },
    {
      "epoch": 0.03549088931193908,
      "grad_norm": 0.478515625,
      "learning_rate": 7.07482993197279e-06,
      "loss": 0.7101,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 261
    },
    {
      "epoch": 0.03562686973075877,
      "grad_norm": 0.361328125,
      "learning_rate": 7.102040816326531e-06,
      "loss": 0.6751,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 262
    },
    {
      "epoch": 0.03576285014957846,
      "grad_norm": 0.359375,
      "learning_rate": 7.129251700680273e-06,
      "loss": 0.68,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 263
    },
    {
      "epoch": 0.03589883056839815,
      "grad_norm": 0.34765625,
      "learning_rate": 7.1564625850340144e-06,
      "loss": 0.7606,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 264
    },
    {
      "epoch": 0.03603481098721784,
      "grad_norm": 0.5625,
      "learning_rate": 7.183673469387755e-06,
      "loss": 0.7028,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 265
    },
    {
      "epoch": 0.03617079140603753,
      "grad_norm": 0.486328125,
      "learning_rate": 7.210884353741497e-06,
      "loss": 0.9393,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 266
    },
    {
      "epoch": 0.03630677182485722,
      "grad_norm": 0.3203125,
      "learning_rate": 7.238095238095239e-06,
      "loss": 0.6763,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 267
    },
    {
      "epoch": 0.036442752243676914,
      "grad_norm": 0.4609375,
      "learning_rate": 7.2653061224489805e-06,
      "loss": 0.7218,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 268
    },
    {
      "epoch": 0.0365787326624966,
      "grad_norm": 0.494140625,
      "learning_rate": 7.292517006802721e-06,
      "loss": 0.7088,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 269
    },
    {
      "epoch": 0.03671471308131629,
      "grad_norm": 0.72265625,
      "learning_rate": 7.319727891156463e-06,
      "loss": 0.8929,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 270
    },
    {
      "epoch": 0.03685069350013598,
      "grad_norm": 0.3515625,
      "learning_rate": 7.346938775510205e-06,
      "loss": 0.6882,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 271
    },
    {
      "epoch": 0.03698667391895567,
      "grad_norm": 0.353515625,
      "learning_rate": 7.374149659863946e-06,
      "loss": 0.5739,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 272
    },
    {
      "epoch": 0.03712265433777536,
      "grad_norm": 0.31640625,
      "learning_rate": 7.4013605442176875e-06,
      "loss": 0.6309,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 273
    },
    {
      "epoch": 0.03725863475659505,
      "grad_norm": 0.43359375,
      "learning_rate": 7.428571428571429e-06,
      "loss": 0.7231,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 274
    },
    {
      "epoch": 0.03739461517541474,
      "grad_norm": 0.333984375,
      "learning_rate": 7.455782312925171e-06,
      "loss": 0.6932,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 275
    },
    {
      "epoch": 0.03753059559423443,
      "grad_norm": 0.4375,
      "learning_rate": 7.482993197278913e-06,
      "loss": 0.7718,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 276
    },
    {
      "epoch": 0.03766657601305412,
      "grad_norm": 0.58984375,
      "learning_rate": 7.5102040816326536e-06,
      "loss": 0.874,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 277
    },
    {
      "epoch": 0.03780255643187381,
      "grad_norm": 0.470703125,
      "learning_rate": 7.537414965986395e-06,
      "loss": 0.6112,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 278
    },
    {
      "epoch": 0.0379385368506935,
      "grad_norm": 0.32421875,
      "learning_rate": 7.564625850340137e-06,
      "loss": 0.7163,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 279
    },
    {
      "epoch": 0.03807451726951319,
      "grad_norm": 0.470703125,
      "learning_rate": 7.591836734693878e-06,
      "loss": 0.7707,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 280
    },
    {
      "epoch": 0.03821049768833288,
      "grad_norm": 0.455078125,
      "learning_rate": 7.61904761904762e-06,
      "loss": 0.6704,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 281
    },
    {
      "epoch": 0.03834647810715257,
      "grad_norm": 0.345703125,
      "learning_rate": 7.646258503401362e-06,
      "loss": 0.5549,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 282
    },
    {
      "epoch": 0.03848245852597226,
      "grad_norm": 0.3203125,
      "learning_rate": 7.673469387755102e-06,
      "loss": 0.6187,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 283
    },
    {
      "epoch": 0.03861843894479195,
      "grad_norm": 0.8125,
      "learning_rate": 7.700680272108844e-06,
      "loss": 0.8107,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 284
    },
    {
      "epoch": 0.03875441936361164,
      "grad_norm": 0.408203125,
      "learning_rate": 7.727891156462586e-06,
      "loss": 0.8127,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 285
    },
    {
      "epoch": 0.03889039978243133,
      "grad_norm": 0.427734375,
      "learning_rate": 7.755102040816327e-06,
      "loss": 0.6056,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 286
    },
    {
      "epoch": 0.03902638020125102,
      "grad_norm": 0.369140625,
      "learning_rate": 7.78231292517007e-06,
      "loss": 0.7949,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 287
    },
    {
      "epoch": 0.03916236062007071,
      "grad_norm": 0.392578125,
      "learning_rate": 7.809523809523811e-06,
      "loss": 0.8029,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 288
    },
    {
      "epoch": 0.0392983410388904,
      "grad_norm": 0.546875,
      "learning_rate": 7.836734693877551e-06,
      "loss": 0.6792,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 289
    },
    {
      "epoch": 0.039434321457710086,
      "grad_norm": 0.5703125,
      "learning_rate": 7.863945578231293e-06,
      "loss": 0.6167,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 290
    },
    {
      "epoch": 0.03957030187652978,
      "grad_norm": 0.55859375,
      "learning_rate": 7.891156462585034e-06,
      "loss": 0.6478,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 291
    },
    {
      "epoch": 0.03970628229534947,
      "grad_norm": 0.40234375,
      "learning_rate": 7.918367346938776e-06,
      "loss": 0.6825,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 292
    },
    {
      "epoch": 0.03984226271416916,
      "grad_norm": 0.427734375,
      "learning_rate": 7.945578231292518e-06,
      "loss": 0.6628,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 293
    },
    {
      "epoch": 0.03997824313298885,
      "grad_norm": 2.09375,
      "learning_rate": 7.97278911564626e-06,
      "loss": 0.9548,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 294
    },
    {
      "epoch": 0.04011422355180854,
      "grad_norm": 0.474609375,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.6634,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 295
    },
    {
      "epoch": 0.04025020397062823,
      "grad_norm": 0.357421875,
      "learning_rate": 8.027210884353741e-06,
      "loss": 0.7087,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 296
    },
    {
      "epoch": 0.04038618438944792,
      "grad_norm": 0.478515625,
      "learning_rate": 8.054421768707483e-06,
      "loss": 0.7059,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 297
    },
    {
      "epoch": 0.04052216480826761,
      "grad_norm": 0.40234375,
      "learning_rate": 8.081632653061225e-06,
      "loss": 0.5822,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 298
    },
    {
      "epoch": 0.0406581452270873,
      "grad_norm": 0.2734375,
      "learning_rate": 8.108843537414967e-06,
      "loss": 0.5086,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 299
    },
    {
      "epoch": 0.04079412564590699,
      "grad_norm": 0.294921875,
      "learning_rate": 8.136054421768708e-06,
      "loss": 0.5651,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 300
    },
    {
      "epoch": 0.040930106064726676,
      "grad_norm": 0.40234375,
      "learning_rate": 8.16326530612245e-06,
      "loss": 0.7184,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 301
    },
    {
      "epoch": 0.04106608648354637,
      "grad_norm": 0.39453125,
      "learning_rate": 8.190476190476192e-06,
      "loss": 0.6715,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 302
    },
    {
      "epoch": 0.04120206690236606,
      "grad_norm": 0.4609375,
      "learning_rate": 8.217687074829933e-06,
      "loss": 0.7961,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 303
    },
    {
      "epoch": 0.04133804732118575,
      "grad_norm": 0.5,
      "learning_rate": 8.244897959183674e-06,
      "loss": 0.7453,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 304
    },
    {
      "epoch": 0.04147402774000544,
      "grad_norm": 0.78125,
      "learning_rate": 8.272108843537415e-06,
      "loss": 0.9214,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 305
    },
    {
      "epoch": 0.04161000815882513,
      "grad_norm": 0.474609375,
      "learning_rate": 8.299319727891157e-06,
      "loss": 0.7075,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 306
    },
    {
      "epoch": 0.04174598857764482,
      "grad_norm": 0.498046875,
      "learning_rate": 8.326530612244899e-06,
      "loss": 0.7676,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 307
    },
    {
      "epoch": 0.04188196899646451,
      "grad_norm": 0.314453125,
      "learning_rate": 8.35374149659864e-06,
      "loss": 0.4648,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 308
    },
    {
      "epoch": 0.0420179494152842,
      "grad_norm": 0.447265625,
      "learning_rate": 8.380952380952382e-06,
      "loss": 0.6919,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 309
    },
    {
      "epoch": 0.04215392983410389,
      "grad_norm": 0.349609375,
      "learning_rate": 8.408163265306122e-06,
      "loss": 0.5938,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 310
    },
    {
      "epoch": 0.04228991025292358,
      "grad_norm": 0.34765625,
      "learning_rate": 8.435374149659866e-06,
      "loss": 0.7843,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 311
    },
    {
      "epoch": 0.042425890671743266,
      "grad_norm": 0.40234375,
      "learning_rate": 8.462585034013606e-06,
      "loss": 0.672,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 312
    },
    {
      "epoch": 0.04256187109056296,
      "grad_norm": 0.275390625,
      "learning_rate": 8.489795918367347e-06,
      "loss": 0.487,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 313
    },
    {
      "epoch": 0.04269785150938265,
      "grad_norm": 0.380859375,
      "learning_rate": 8.517006802721089e-06,
      "loss": 0.8018,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 314
    },
    {
      "epoch": 0.04283383192820234,
      "grad_norm": 0.431640625,
      "learning_rate": 8.544217687074831e-06,
      "loss": 0.6273,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 315
    },
    {
      "epoch": 0.04296981234702203,
      "grad_norm": 0.345703125,
      "learning_rate": 8.571428571428571e-06,
      "loss": 0.6175,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 316
    },
    {
      "epoch": 0.04310579276584172,
      "grad_norm": 1.453125,
      "learning_rate": 8.598639455782314e-06,
      "loss": 0.8319,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 317
    },
    {
      "epoch": 0.04324177318466141,
      "grad_norm": 0.322265625,
      "learning_rate": 8.625850340136054e-06,
      "loss": 0.52,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 318
    },
    {
      "epoch": 0.043377753603481096,
      "grad_norm": 1.125,
      "learning_rate": 8.653061224489798e-06,
      "loss": 0.5791,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 319
    },
    {
      "epoch": 0.04351373402230079,
      "grad_norm": 0.470703125,
      "learning_rate": 8.680272108843538e-06,
      "loss": 0.7443,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 320
    },
    {
      "epoch": 0.04364971444112048,
      "grad_norm": 0.400390625,
      "learning_rate": 8.70748299319728e-06,
      "loss": 0.8138,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 321
    },
    {
      "epoch": 0.04378569485994017,
      "grad_norm": 0.404296875,
      "learning_rate": 8.734693877551021e-06,
      "loss": 0.7811,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 322
    },
    {
      "epoch": 0.043921675278759856,
      "grad_norm": 0.3828125,
      "learning_rate": 8.761904761904763e-06,
      "loss": 0.7497,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 323
    },
    {
      "epoch": 0.04405765569757955,
      "grad_norm": 0.314453125,
      "learning_rate": 8.789115646258503e-06,
      "loss": 0.5591,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 324
    },
    {
      "epoch": 0.04419363611639924,
      "grad_norm": 0.75390625,
      "learning_rate": 8.816326530612247e-06,
      "loss": 0.8439,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 325
    },
    {
      "epoch": 0.04432961653521893,
      "grad_norm": 0.3515625,
      "learning_rate": 8.843537414965987e-06,
      "loss": 0.6481,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 326
    },
    {
      "epoch": 0.044465596954038615,
      "grad_norm": 0.5625,
      "learning_rate": 8.87074829931973e-06,
      "loss": 0.745,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 327
    },
    {
      "epoch": 0.04460157737285831,
      "grad_norm": 0.40625,
      "learning_rate": 8.89795918367347e-06,
      "loss": 0.4888,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 328
    },
    {
      "epoch": 0.044737557791678,
      "grad_norm": 0.55078125,
      "learning_rate": 8.925170068027212e-06,
      "loss": 0.971,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 329
    },
    {
      "epoch": 0.044873538210497686,
      "grad_norm": 0.408203125,
      "learning_rate": 8.952380952380953e-06,
      "loss": 0.8376,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 330
    },
    {
      "epoch": 0.04500951862931738,
      "grad_norm": 0.326171875,
      "learning_rate": 8.979591836734695e-06,
      "loss": 0.5605,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 331
    },
    {
      "epoch": 0.04514549904813707,
      "grad_norm": 0.341796875,
      "learning_rate": 9.006802721088435e-06,
      "loss": 0.5184,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 332
    },
    {
      "epoch": 0.04528147946695676,
      "grad_norm": 0.3515625,
      "learning_rate": 9.034013605442179e-06,
      "loss": 0.6974,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 333
    },
    {
      "epoch": 0.045417459885776446,
      "grad_norm": 0.333984375,
      "learning_rate": 9.061224489795919e-06,
      "loss": 0.6138,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 334
    },
    {
      "epoch": 0.04555344030459614,
      "grad_norm": 0.365234375,
      "learning_rate": 9.08843537414966e-06,
      "loss": 0.6425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 335
    },
    {
      "epoch": 0.04568942072341583,
      "grad_norm": 0.306640625,
      "learning_rate": 9.115646258503402e-06,
      "loss": 0.5353,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 336
    },
    {
      "epoch": 0.04582540114223552,
      "grad_norm": 0.49609375,
      "learning_rate": 9.142857142857144e-06,
      "loss": 0.7184,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 337
    },
    {
      "epoch": 0.045961381561055205,
      "grad_norm": 0.62890625,
      "learning_rate": 9.170068027210886e-06,
      "loss": 0.8455,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 338
    },
    {
      "epoch": 0.0460973619798749,
      "grad_norm": 0.380859375,
      "learning_rate": 9.197278911564627e-06,
      "loss": 0.7378,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 339
    },
    {
      "epoch": 0.04623334239869459,
      "grad_norm": 0.671875,
      "learning_rate": 9.224489795918367e-06,
      "loss": 0.7215,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 340
    },
    {
      "epoch": 0.046369322817514276,
      "grad_norm": 0.369140625,
      "learning_rate": 9.251700680272109e-06,
      "loss": 0.7313,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 341
    },
    {
      "epoch": 0.04650530323633397,
      "grad_norm": 0.34765625,
      "learning_rate": 9.278911564625851e-06,
      "loss": 0.6239,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 342
    },
    {
      "epoch": 0.04664128365515366,
      "grad_norm": 0.34375,
      "learning_rate": 9.306122448979593e-06,
      "loss": 0.5005,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 343
    },
    {
      "epoch": 0.04677726407397335,
      "grad_norm": 0.427734375,
      "learning_rate": 9.333333333333334e-06,
      "loss": 0.8174,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 344
    },
    {
      "epoch": 0.046913244492793035,
      "grad_norm": 0.4453125,
      "learning_rate": 9.360544217687076e-06,
      "loss": 0.7788,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 345
    },
    {
      "epoch": 0.04704922491161273,
      "grad_norm": 0.310546875,
      "learning_rate": 9.387755102040818e-06,
      "loss": 0.6099,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 346
    },
    {
      "epoch": 0.04718520533043242,
      "grad_norm": 0.482421875,
      "learning_rate": 9.414965986394558e-06,
      "loss": 0.6987,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 347
    },
    {
      "epoch": 0.047321185749252107,
      "grad_norm": 0.4296875,
      "learning_rate": 9.4421768707483e-06,
      "loss": 0.7588,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 348
    },
    {
      "epoch": 0.047457166168071795,
      "grad_norm": 0.6953125,
      "learning_rate": 9.469387755102041e-06,
      "loss": 0.856,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 349
    },
    {
      "epoch": 0.04759314658689149,
      "grad_norm": 0.67578125,
      "learning_rate": 9.496598639455783e-06,
      "loss": 0.9992,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 350
    },
    {
      "epoch": 0.04772912700571118,
      "grad_norm": 0.43359375,
      "learning_rate": 9.523809523809525e-06,
      "loss": 0.7578,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 351
    },
    {
      "epoch": 0.047865107424530866,
      "grad_norm": 0.2119140625,
      "learning_rate": 9.551020408163266e-06,
      "loss": 0.3485,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 352
    },
    {
      "epoch": 0.04800108784335056,
      "grad_norm": 0.443359375,
      "learning_rate": 9.578231292517007e-06,
      "loss": 0.658,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 353
    },
    {
      "epoch": 0.04813706826217025,
      "grad_norm": 0.416015625,
      "learning_rate": 9.60544217687075e-06,
      "loss": 0.8817,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 354
    },
    {
      "epoch": 0.04827304868098994,
      "grad_norm": 0.423828125,
      "learning_rate": 9.63265306122449e-06,
      "loss": 0.6756,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 355
    },
    {
      "epoch": 0.048409029099809625,
      "grad_norm": 0.3515625,
      "learning_rate": 9.659863945578232e-06,
      "loss": 0.7754,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 356
    },
    {
      "epoch": 0.04854500951862932,
      "grad_norm": 0.4921875,
      "learning_rate": 9.687074829931973e-06,
      "loss": 0.8369,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 357
    },
    {
      "epoch": 0.04868098993744901,
      "grad_norm": 0.353515625,
      "learning_rate": 9.714285714285715e-06,
      "loss": 0.5322,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 358
    },
    {
      "epoch": 0.048816970356268696,
      "grad_norm": 0.49609375,
      "learning_rate": 9.741496598639457e-06,
      "loss": 0.9388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 359
    },
    {
      "epoch": 0.048952950775088384,
      "grad_norm": 0.3515625,
      "learning_rate": 9.768707482993199e-06,
      "loss": 0.6512,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 360
    },
    {
      "epoch": 0.04908893119390808,
      "grad_norm": 0.25390625,
      "learning_rate": 9.795918367346939e-06,
      "loss": 0.4749,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 361
    },
    {
      "epoch": 0.04922491161272777,
      "grad_norm": 0.578125,
      "learning_rate": 9.823129251700682e-06,
      "loss": 0.7264,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 362
    },
    {
      "epoch": 0.049360892031547456,
      "grad_norm": 0.5234375,
      "learning_rate": 9.850340136054422e-06,
      "loss": 0.5392,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 363
    },
    {
      "epoch": 0.049496872450367144,
      "grad_norm": 0.44921875,
      "learning_rate": 9.877551020408164e-06,
      "loss": 0.6844,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 364
    },
    {
      "epoch": 0.04963285286918684,
      "grad_norm": 0.2412109375,
      "learning_rate": 9.904761904761906e-06,
      "loss": 0.4134,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 365
    },
    {
      "epoch": 0.04976883328800653,
      "grad_norm": 0.345703125,
      "learning_rate": 9.931972789115647e-06,
      "loss": 0.6761,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 366
    },
    {
      "epoch": 0.049904813706826215,
      "grad_norm": 0.474609375,
      "learning_rate": 9.959183673469387e-06,
      "loss": 0.6971,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 367
    },
    {
      "epoch": 0.05004079412564591,
      "grad_norm": 0.275390625,
      "learning_rate": 9.98639455782313e-06,
      "loss": 0.4684,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 368
    },
    {
      "epoch": 0.0501767745444656,
      "grad_norm": 0.37109375,
      "learning_rate": 1.0013605442176873e-05,
      "loss": 0.6971,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 369
    },
    {
      "epoch": 0.050312754963285286,
      "grad_norm": 0.375,
      "learning_rate": 1.0040816326530614e-05,
      "loss": 0.5433,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 370
    },
    {
      "epoch": 0.050448735382104974,
      "grad_norm": 0.5390625,
      "learning_rate": 1.0068027210884354e-05,
      "loss": 0.8608,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 371
    },
    {
      "epoch": 0.05058471580092467,
      "grad_norm": 0.3125,
      "learning_rate": 1.0095238095238096e-05,
      "loss": 0.6313,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 372
    },
    {
      "epoch": 0.05072069621974436,
      "grad_norm": 0.35546875,
      "learning_rate": 1.0122448979591836e-05,
      "loss": 0.5334,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 373
    },
    {
      "epoch": 0.050856676638564045,
      "grad_norm": 0.474609375,
      "learning_rate": 1.014965986394558e-05,
      "loss": 0.9106,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 374
    },
    {
      "epoch": 0.05099265705738373,
      "grad_norm": 0.50390625,
      "learning_rate": 1.0176870748299321e-05,
      "loss": 0.802,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 375
    },
    {
      "epoch": 0.05112863747620343,
      "grad_norm": 0.400390625,
      "learning_rate": 1.0204081632653063e-05,
      "loss": 0.5729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 376
    },
    {
      "epoch": 0.05126461789502312,
      "grad_norm": 0.328125,
      "learning_rate": 1.0231292517006803e-05,
      "loss": 0.5594,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 377
    },
    {
      "epoch": 0.051400598313842805,
      "grad_norm": 0.40234375,
      "learning_rate": 1.0258503401360545e-05,
      "loss": 0.7295,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 378
    },
    {
      "epoch": 0.0515365787326625,
      "grad_norm": 0.498046875,
      "learning_rate": 1.0285714285714285e-05,
      "loss": 0.8379,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 379
    },
    {
      "epoch": 0.05167255915148219,
      "grad_norm": 0.60546875,
      "learning_rate": 1.0312925170068028e-05,
      "loss": 0.7363,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 380
    },
    {
      "epoch": 0.051808539570301876,
      "grad_norm": 0.84765625,
      "learning_rate": 1.034013605442177e-05,
      "loss": 0.7749,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 381
    },
    {
      "epoch": 0.051944519989121564,
      "grad_norm": 0.283203125,
      "learning_rate": 1.0367346938775512e-05,
      "loss": 0.4989,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 382
    },
    {
      "epoch": 0.05208050040794126,
      "grad_norm": 0.330078125,
      "learning_rate": 1.0394557823129252e-05,
      "loss": 0.5752,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 383
    },
    {
      "epoch": 0.05221648082676095,
      "grad_norm": 0.263671875,
      "learning_rate": 1.0421768707482993e-05,
      "loss": 0.4671,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 384
    },
    {
      "epoch": 0.052352461245580635,
      "grad_norm": 0.466796875,
      "learning_rate": 1.0448979591836737e-05,
      "loss": 0.7692,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 385
    },
    {
      "epoch": 0.05248844166440032,
      "grad_norm": 0.390625,
      "learning_rate": 1.0476190476190477e-05,
      "loss": 0.495,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 386
    },
    {
      "epoch": 0.05262442208322002,
      "grad_norm": 0.4609375,
      "learning_rate": 1.0503401360544219e-05,
      "loss": 0.6807,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 387
    },
    {
      "epoch": 0.052760402502039706,
      "grad_norm": 0.412109375,
      "learning_rate": 1.053061224489796e-05,
      "loss": 0.6576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 388
    },
    {
      "epoch": 0.052896382920859394,
      "grad_norm": 0.333984375,
      "learning_rate": 1.05578231292517e-05,
      "loss": 0.5825,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 389
    },
    {
      "epoch": 0.05303236333967909,
      "grad_norm": 0.478515625,
      "learning_rate": 1.0585034013605444e-05,
      "loss": 0.8063,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 390
    },
    {
      "epoch": 0.05316834375849878,
      "grad_norm": 0.263671875,
      "learning_rate": 1.0612244897959186e-05,
      "loss": 0.5661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 391
    },
    {
      "epoch": 0.053304324177318466,
      "grad_norm": 0.5234375,
      "learning_rate": 1.0639455782312926e-05,
      "loss": 0.9085,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 392
    },
    {
      "epoch": 0.053440304596138154,
      "grad_norm": 0.65625,
      "learning_rate": 1.0666666666666667e-05,
      "loss": 0.8589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 393
    },
    {
      "epoch": 0.05357628501495785,
      "grad_norm": 0.4453125,
      "learning_rate": 1.0693877551020409e-05,
      "loss": 0.7983,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 394
    },
    {
      "epoch": 0.05371226543377754,
      "grad_norm": 0.5859375,
      "learning_rate": 1.0721088435374149e-05,
      "loss": 0.9154,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 395
    },
    {
      "epoch": 0.053848245852597225,
      "grad_norm": 0.54296875,
      "learning_rate": 1.0748299319727893e-05,
      "loss": 0.7284,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 396
    },
    {
      "epoch": 0.05398422627141691,
      "grad_norm": 0.287109375,
      "learning_rate": 1.0775510204081634e-05,
      "loss": 0.5695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 397
    },
    {
      "epoch": 0.05412020669023661,
      "grad_norm": 0.53125,
      "learning_rate": 1.0802721088435374e-05,
      "loss": 0.6257,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 398
    },
    {
      "epoch": 0.054256187109056296,
      "grad_norm": 0.5390625,
      "learning_rate": 1.0829931972789116e-05,
      "loss": 0.8249,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 399
    },
    {
      "epoch": 0.054392167527875984,
      "grad_norm": 0.380859375,
      "learning_rate": 1.0857142857142858e-05,
      "loss": 0.6406,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 400
    },
    {
      "epoch": 0.05452814794669568,
      "grad_norm": 0.353515625,
      "learning_rate": 1.0884353741496601e-05,
      "loss": 0.6986,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 401
    },
    {
      "epoch": 0.05466412836551537,
      "grad_norm": 0.451171875,
      "learning_rate": 1.0911564625850341e-05,
      "loss": 0.7218,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 402
    },
    {
      "epoch": 0.054800108784335055,
      "grad_norm": 0.25390625,
      "learning_rate": 1.0938775510204083e-05,
      "loss": 0.4718,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 403
    },
    {
      "epoch": 0.054936089203154743,
      "grad_norm": 1.015625,
      "learning_rate": 1.0965986394557823e-05,
      "loss": 0.9548,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 404
    },
    {
      "epoch": 0.05507206962197444,
      "grad_norm": 0.453125,
      "learning_rate": 1.0993197278911565e-05,
      "loss": 0.5225,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 405
    },
    {
      "epoch": 0.05520805004079413,
      "grad_norm": 0.8125,
      "learning_rate": 1.1020408163265306e-05,
      "loss": 0.6984,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 406
    },
    {
      "epoch": 0.055344030459613815,
      "grad_norm": 0.337890625,
      "learning_rate": 1.104761904761905e-05,
      "loss": 0.6056,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 407
    },
    {
      "epoch": 0.0554800108784335,
      "grad_norm": 0.3203125,
      "learning_rate": 1.107482993197279e-05,
      "loss": 0.5592,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 408
    },
    {
      "epoch": 0.0556159912972532,
      "grad_norm": 0.50390625,
      "learning_rate": 1.1102040816326532e-05,
      "loss": 0.7472,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 409
    },
    {
      "epoch": 0.055751971716072886,
      "grad_norm": 0.51953125,
      "learning_rate": 1.1129251700680272e-05,
      "loss": 0.8509,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 410
    },
    {
      "epoch": 0.055887952134892574,
      "grad_norm": 0.416015625,
      "learning_rate": 1.1156462585034013e-05,
      "loss": 0.6698,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 411
    },
    {
      "epoch": 0.05602393255371226,
      "grad_norm": 0.453125,
      "learning_rate": 1.1183673469387757e-05,
      "loss": 0.5566,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 412
    },
    {
      "epoch": 0.05615991297253196,
      "grad_norm": 0.37890625,
      "learning_rate": 1.1210884353741499e-05,
      "loss": 0.7792,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 413
    },
    {
      "epoch": 0.056295893391351645,
      "grad_norm": 0.609375,
      "learning_rate": 1.1238095238095239e-05,
      "loss": 0.5747,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 414
    },
    {
      "epoch": 0.05643187381017133,
      "grad_norm": 0.43359375,
      "learning_rate": 1.126530612244898e-05,
      "loss": 0.8993,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 415
    },
    {
      "epoch": 0.05656785422899103,
      "grad_norm": 0.5703125,
      "learning_rate": 1.1292517006802722e-05,
      "loss": 0.7666,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 416
    },
    {
      "epoch": 0.056703834647810716,
      "grad_norm": 0.72265625,
      "learning_rate": 1.1319727891156464e-05,
      "loss": 0.5433,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 417
    },
    {
      "epoch": 0.056839815066630404,
      "grad_norm": 0.388671875,
      "learning_rate": 1.1346938775510206e-05,
      "loss": 0.7432,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 418
    },
    {
      "epoch": 0.05697579548545009,
      "grad_norm": 0.34765625,
      "learning_rate": 1.1374149659863947e-05,
      "loss": 0.7215,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 419
    },
    {
      "epoch": 0.05711177590426979,
      "grad_norm": 0.484375,
      "learning_rate": 1.1401360544217687e-05,
      "loss": 0.9536,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 420
    },
    {
      "epoch": 0.057247756323089476,
      "grad_norm": 0.62890625,
      "learning_rate": 1.1428571428571429e-05,
      "loss": 0.9642,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 421
    },
    {
      "epoch": 0.057383736741909164,
      "grad_norm": 0.546875,
      "learning_rate": 1.145578231292517e-05,
      "loss": 0.7106,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 422
    },
    {
      "epoch": 0.05751971716072885,
      "grad_norm": 0.953125,
      "learning_rate": 1.1482993197278912e-05,
      "loss": 1.1024,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 423
    },
    {
      "epoch": 0.05765569757954855,
      "grad_norm": 0.40625,
      "learning_rate": 1.1510204081632654e-05,
      "loss": 0.5381,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 424
    },
    {
      "epoch": 0.057791677998368235,
      "grad_norm": 0.361328125,
      "learning_rate": 1.1537414965986396e-05,
      "loss": 0.6531,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 425
    },
    {
      "epoch": 0.05792765841718792,
      "grad_norm": 0.435546875,
      "learning_rate": 1.1564625850340136e-05,
      "loss": 0.5202,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 426
    },
    {
      "epoch": 0.05806363883600762,
      "grad_norm": 0.82421875,
      "learning_rate": 1.1591836734693878e-05,
      "loss": 0.6372,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 427
    },
    {
      "epoch": 0.058199619254827306,
      "grad_norm": 0.404296875,
      "learning_rate": 1.1619047619047621e-05,
      "loss": 0.5677,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 428
    },
    {
      "epoch": 0.058335599673646994,
      "grad_norm": 0.419921875,
      "learning_rate": 1.1646258503401361e-05,
      "loss": 0.7122,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 429
    },
    {
      "epoch": 0.05847158009246668,
      "grad_norm": 0.408203125,
      "learning_rate": 1.1673469387755103e-05,
      "loss": 0.7005,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 430
    },
    {
      "epoch": 0.05860756051128638,
      "grad_norm": 0.435546875,
      "learning_rate": 1.1700680272108845e-05,
      "loss": 0.665,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 431
    },
    {
      "epoch": 0.058743540930106065,
      "grad_norm": 0.345703125,
      "learning_rate": 1.1727891156462585e-05,
      "loss": 0.5583,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 432
    },
    {
      "epoch": 0.058879521348925754,
      "grad_norm": 0.69140625,
      "learning_rate": 1.1755102040816328e-05,
      "loss": 0.8818,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 433
    },
    {
      "epoch": 0.05901550176774544,
      "grad_norm": 0.67578125,
      "learning_rate": 1.178231292517007e-05,
      "loss": 0.8172,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 434
    },
    {
      "epoch": 0.05915148218656514,
      "grad_norm": 0.26171875,
      "learning_rate": 1.180952380952381e-05,
      "loss": 0.5228,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 435
    },
    {
      "epoch": 0.059287462605384825,
      "grad_norm": 0.34765625,
      "learning_rate": 1.1836734693877552e-05,
      "loss": 0.6009,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 436
    },
    {
      "epoch": 0.05942344302420451,
      "grad_norm": 0.58984375,
      "learning_rate": 1.1863945578231293e-05,
      "loss": 0.6424,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 437
    },
    {
      "epoch": 0.05955942344302421,
      "grad_norm": 0.51171875,
      "learning_rate": 1.1891156462585033e-05,
      "loss": 0.9295,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 438
    },
    {
      "epoch": 0.059695403861843896,
      "grad_norm": 0.37890625,
      "learning_rate": 1.1918367346938777e-05,
      "loss": 0.7046,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 439
    },
    {
      "epoch": 0.059831384280663584,
      "grad_norm": 0.400390625,
      "learning_rate": 1.1945578231292519e-05,
      "loss": 0.6343,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 440
    },
    {
      "epoch": 0.05996736469948327,
      "grad_norm": 0.53125,
      "learning_rate": 1.197278911564626e-05,
      "loss": 0.8281,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 441
    },
    {
      "epoch": 0.06010334511830297,
      "grad_norm": 0.462890625,
      "learning_rate": 1.2e-05,
      "loss": 0.7962,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 442
    },
    {
      "epoch": 0.060239325537122655,
      "grad_norm": 0.34375,
      "learning_rate": 1.2027210884353742e-05,
      "loss": 0.5674,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 443
    },
    {
      "epoch": 0.06037530595594234,
      "grad_norm": 0.32421875,
      "learning_rate": 1.2054421768707485e-05,
      "loss": 0.5415,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 444
    },
    {
      "epoch": 0.06051128637476203,
      "grad_norm": 0.306640625,
      "learning_rate": 1.2081632653061225e-05,
      "loss": 0.505,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 445
    },
    {
      "epoch": 0.060647266793581726,
      "grad_norm": 0.55078125,
      "learning_rate": 1.2108843537414967e-05,
      "loss": 0.783,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 446
    },
    {
      "epoch": 0.060783247212401414,
      "grad_norm": 0.44140625,
      "learning_rate": 1.2136054421768709e-05,
      "loss": 0.613,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 447
    },
    {
      "epoch": 0.0609192276312211,
      "grad_norm": 3.375,
      "learning_rate": 1.2163265306122449e-05,
      "loss": 0.8394,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 448
    },
    {
      "epoch": 0.06105520805004079,
      "grad_norm": 0.6796875,
      "learning_rate": 1.2190476190476192e-05,
      "loss": 0.7384,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 449
    },
    {
      "epoch": 0.061191188468860486,
      "grad_norm": 0.291015625,
      "learning_rate": 1.2217687074829934e-05,
      "loss": 0.5768,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 450
    },
    {
      "epoch": 0.061327168887680174,
      "grad_norm": 0.4296875,
      "learning_rate": 1.2244897959183674e-05,
      "loss": 0.7186,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 451
    },
    {
      "epoch": 0.06146314930649986,
      "grad_norm": 0.5703125,
      "learning_rate": 1.2272108843537416e-05,
      "loss": 0.9295,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 452
    },
    {
      "epoch": 0.06159912972531956,
      "grad_norm": 0.443359375,
      "learning_rate": 1.2299319727891158e-05,
      "loss": 0.6816,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 453
    },
    {
      "epoch": 0.061735110144139245,
      "grad_norm": 0.59765625,
      "learning_rate": 1.2326530612244898e-05,
      "loss": 0.826,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 454
    },
    {
      "epoch": 0.06187109056295893,
      "grad_norm": 0.6328125,
      "learning_rate": 1.2353741496598641e-05,
      "loss": 0.6152,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 455
    },
    {
      "epoch": 0.06200707098177862,
      "grad_norm": 0.60546875,
      "learning_rate": 1.2380952380952383e-05,
      "loss": 0.5968,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 456
    },
    {
      "epoch": 0.062143051400598316,
      "grad_norm": 0.4609375,
      "learning_rate": 1.2408163265306123e-05,
      "loss": 0.6663,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 457
    },
    {
      "epoch": 0.062279031819418004,
      "grad_norm": 0.3359375,
      "learning_rate": 1.2435374149659865e-05,
      "loss": 0.5983,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 458
    },
    {
      "epoch": 0.06241501223823769,
      "grad_norm": 0.515625,
      "learning_rate": 1.2462585034013606e-05,
      "loss": 0.8436,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 459
    },
    {
      "epoch": 0.06255099265705738,
      "grad_norm": 0.6640625,
      "learning_rate": 1.248979591836735e-05,
      "loss": 0.6678,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 460
    },
    {
      "epoch": 0.06268697307587708,
      "grad_norm": 1.2109375,
      "learning_rate": 1.251700680272109e-05,
      "loss": 1.0929,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 461
    },
    {
      "epoch": 0.06282295349469677,
      "grad_norm": 0.482421875,
      "learning_rate": 1.2544217687074832e-05,
      "loss": 0.7142,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 462
    },
    {
      "epoch": 0.06295893391351645,
      "grad_norm": 0.466796875,
      "learning_rate": 1.2571428571428572e-05,
      "loss": 0.6984,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 463
    },
    {
      "epoch": 0.06309491433233615,
      "grad_norm": 0.2353515625,
      "learning_rate": 1.2598639455782313e-05,
      "loss": 0.4098,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 464
    },
    {
      "epoch": 0.06323089475115583,
      "grad_norm": 0.365234375,
      "learning_rate": 1.2625850340136055e-05,
      "loss": 0.5765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 465
    },
    {
      "epoch": 0.06336687516997552,
      "grad_norm": 0.63671875,
      "learning_rate": 1.2653061224489798e-05,
      "loss": 0.7435,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 466
    },
    {
      "epoch": 0.06350285558879522,
      "grad_norm": 0.376953125,
      "learning_rate": 1.2680272108843539e-05,
      "loss": 0.6115,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 467
    },
    {
      "epoch": 0.0636388360076149,
      "grad_norm": 0.58984375,
      "learning_rate": 1.270748299319728e-05,
      "loss": 0.8625,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 468
    },
    {
      "epoch": 0.0637748164264346,
      "grad_norm": 0.443359375,
      "learning_rate": 1.273469387755102e-05,
      "loss": 0.8615,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 469
    },
    {
      "epoch": 0.06391079684525429,
      "grad_norm": 0.345703125,
      "learning_rate": 1.2761904761904762e-05,
      "loss": 0.6159,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 470
    },
    {
      "epoch": 0.06404677726407397,
      "grad_norm": 0.4375,
      "learning_rate": 1.2789115646258505e-05,
      "loss": 0.7295,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 471
    },
    {
      "epoch": 0.06418275768289367,
      "grad_norm": 0.984375,
      "learning_rate": 1.2816326530612247e-05,
      "loss": 0.9731,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 472
    },
    {
      "epoch": 0.06431873810171335,
      "grad_norm": 0.384765625,
      "learning_rate": 1.2843537414965987e-05,
      "loss": 0.6507,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 473
    },
    {
      "epoch": 0.06445471852053304,
      "grad_norm": 0.279296875,
      "learning_rate": 1.2870748299319729e-05,
      "loss": 0.4395,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 474
    },
    {
      "epoch": 0.06459069893935274,
      "grad_norm": 0.42578125,
      "learning_rate": 1.2897959183673469e-05,
      "loss": 0.7064,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 475
    },
    {
      "epoch": 0.06472667935817242,
      "grad_norm": 0.6796875,
      "learning_rate": 1.2925170068027212e-05,
      "loss": 0.7272,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 476
    },
    {
      "epoch": 0.06486265977699211,
      "grad_norm": 0.5390625,
      "learning_rate": 1.2952380952380954e-05,
      "loss": 0.4587,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 477
    },
    {
      "epoch": 0.06499864019581181,
      "grad_norm": 0.2734375,
      "learning_rate": 1.2979591836734696e-05,
      "loss": 0.564,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 478
    },
    {
      "epoch": 0.06513462061463149,
      "grad_norm": 0.361328125,
      "learning_rate": 1.3006802721088436e-05,
      "loss": 0.5409,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 479
    },
    {
      "epoch": 0.06527060103345118,
      "grad_norm": 0.58984375,
      "learning_rate": 1.3034013605442178e-05,
      "loss": 0.6755,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 480
    },
    {
      "epoch": 0.06540658145227088,
      "grad_norm": 0.462890625,
      "learning_rate": 1.3061224489795918e-05,
      "loss": 0.7492,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 481
    },
    {
      "epoch": 0.06554256187109056,
      "grad_norm": 0.486328125,
      "learning_rate": 1.3088435374149661e-05,
      "loss": 0.8991,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 482
    },
    {
      "epoch": 0.06567854228991026,
      "grad_norm": 0.310546875,
      "learning_rate": 1.3115646258503403e-05,
      "loss": 0.5011,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 483
    },
    {
      "epoch": 0.06581452270872994,
      "grad_norm": 0.232421875,
      "learning_rate": 1.3142857142857145e-05,
      "loss": 0.3862,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 484
    },
    {
      "epoch": 0.06595050312754963,
      "grad_norm": 0.34375,
      "learning_rate": 1.3170068027210885e-05,
      "loss": 0.6554,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 485
    },
    {
      "epoch": 0.06608648354636933,
      "grad_norm": 0.46875,
      "learning_rate": 1.3197278911564626e-05,
      "loss": 0.6403,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 486
    },
    {
      "epoch": 0.06622246396518901,
      "grad_norm": 0.375,
      "learning_rate": 1.322448979591837e-05,
      "loss": 0.7493,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 487
    },
    {
      "epoch": 0.0663584443840087,
      "grad_norm": 0.54296875,
      "learning_rate": 1.325170068027211e-05,
      "loss": 0.6587,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 488
    },
    {
      "epoch": 0.0664944248028284,
      "grad_norm": 0.66796875,
      "learning_rate": 1.3278911564625852e-05,
      "loss": 0.3784,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 489
    },
    {
      "epoch": 0.06663040522164808,
      "grad_norm": 0.6796875,
      "learning_rate": 1.3306122448979593e-05,
      "loss": 0.8104,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 490
    },
    {
      "epoch": 0.06676638564046777,
      "grad_norm": 0.373046875,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.6315,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 491
    },
    {
      "epoch": 0.06690236605928747,
      "grad_norm": 0.421875,
      "learning_rate": 1.3360544217687077e-05,
      "loss": 0.5985,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 492
    },
    {
      "epoch": 0.06703834647810715,
      "grad_norm": 0.431640625,
      "learning_rate": 1.3387755102040818e-05,
      "loss": 0.5402,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 493
    },
    {
      "epoch": 0.06717432689692684,
      "grad_norm": 0.361328125,
      "learning_rate": 1.3414965986394558e-05,
      "loss": 0.6382,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 494
    },
    {
      "epoch": 0.06731030731574653,
      "grad_norm": 0.474609375,
      "learning_rate": 1.34421768707483e-05,
      "loss": 0.8337,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 495
    },
    {
      "epoch": 0.06744628773456622,
      "grad_norm": 0.41796875,
      "learning_rate": 1.3469387755102042e-05,
      "loss": 0.7397,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 496
    },
    {
      "epoch": 0.06758226815338592,
      "grad_norm": 0.34765625,
      "learning_rate": 1.3496598639455782e-05,
      "loss": 0.5174,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 497
    },
    {
      "epoch": 0.0677182485722056,
      "grad_norm": 0.5234375,
      "learning_rate": 1.3523809523809525e-05,
      "loss": 1.0026,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 498
    },
    {
      "epoch": 0.06785422899102529,
      "grad_norm": 0.59375,
      "learning_rate": 1.3551020408163267e-05,
      "loss": 0.7067,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 499
    },
    {
      "epoch": 0.06799020940984499,
      "grad_norm": 0.314453125,
      "learning_rate": 1.3578231292517007e-05,
      "loss": 0.5436,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 500
    },
    {
      "epoch": 0.06812618982866467,
      "grad_norm": 0.31640625,
      "learning_rate": 1.3605442176870749e-05,
      "loss": 0.5347,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 501
    },
    {
      "epoch": 0.06826217024748436,
      "grad_norm": 0.50390625,
      "learning_rate": 1.363265306122449e-05,
      "loss": 0.7526,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 502
    },
    {
      "epoch": 0.06839815066630406,
      "grad_norm": 0.341796875,
      "learning_rate": 1.3659863945578234e-05,
      "loss": 0.6105,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 503
    },
    {
      "epoch": 0.06853413108512374,
      "grad_norm": 0.3515625,
      "learning_rate": 1.3687074829931974e-05,
      "loss": 0.6859,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 504
    },
    {
      "epoch": 0.06867011150394343,
      "grad_norm": 0.33203125,
      "learning_rate": 1.3714285714285716e-05,
      "loss": 0.7122,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 505
    },
    {
      "epoch": 0.06880609192276312,
      "grad_norm": 0.470703125,
      "learning_rate": 1.3741496598639456e-05,
      "loss": 0.8667,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 506
    },
    {
      "epoch": 0.06894207234158281,
      "grad_norm": 0.73046875,
      "learning_rate": 1.3768707482993198e-05,
      "loss": 0.7036,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 507
    },
    {
      "epoch": 0.0690780527604025,
      "grad_norm": 0.455078125,
      "learning_rate": 1.3795918367346941e-05,
      "loss": 0.7152,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 508
    },
    {
      "epoch": 0.06921403317922219,
      "grad_norm": 0.734375,
      "learning_rate": 1.3823129251700683e-05,
      "loss": 0.8081,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 509
    },
    {
      "epoch": 0.06935001359804188,
      "grad_norm": 0.330078125,
      "learning_rate": 1.3850340136054423e-05,
      "loss": 0.5981,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 510
    },
    {
      "epoch": 0.06948599401686158,
      "grad_norm": 3.0625,
      "learning_rate": 1.3877551020408165e-05,
      "loss": 0.7912,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 511
    },
    {
      "epoch": 0.06962197443568126,
      "grad_norm": 0.498046875,
      "learning_rate": 1.3904761904761905e-05,
      "loss": 0.8778,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 512
    },
    {
      "epoch": 0.06975795485450095,
      "grad_norm": 0.38671875,
      "learning_rate": 1.3931972789115646e-05,
      "loss": 0.5428,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 513
    },
    {
      "epoch": 0.06989393527332065,
      "grad_norm": 0.357421875,
      "learning_rate": 1.395918367346939e-05,
      "loss": 0.7606,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 514
    },
    {
      "epoch": 0.07002991569214033,
      "grad_norm": 0.443359375,
      "learning_rate": 1.3986394557823131e-05,
      "loss": 0.5864,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 515
    },
    {
      "epoch": 0.07016589611096002,
      "grad_norm": 0.369140625,
      "learning_rate": 1.4013605442176872e-05,
      "loss": 0.6125,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 516
    },
    {
      "epoch": 0.0703018765297797,
      "grad_norm": 0.416015625,
      "learning_rate": 1.4040816326530613e-05,
      "loss": 0.5894,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 517
    },
    {
      "epoch": 0.0704378569485994,
      "grad_norm": 0.78125,
      "learning_rate": 1.4068027210884353e-05,
      "loss": 0.8887,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 518
    },
    {
      "epoch": 0.0705738373674191,
      "grad_norm": 0.423828125,
      "learning_rate": 1.4095238095238097e-05,
      "loss": 0.6815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 519
    },
    {
      "epoch": 0.07070981778623878,
      "grad_norm": 0.41796875,
      "learning_rate": 1.4122448979591838e-05,
      "loss": 0.8229,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 520
    },
    {
      "epoch": 0.07084579820505847,
      "grad_norm": 0.68359375,
      "learning_rate": 1.414965986394558e-05,
      "loss": 0.7354,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 521
    },
    {
      "epoch": 0.07098177862387817,
      "grad_norm": 0.515625,
      "learning_rate": 1.417687074829932e-05,
      "loss": 0.5526,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 522
    },
    {
      "epoch": 0.07111775904269785,
      "grad_norm": 0.423828125,
      "learning_rate": 1.4204081632653062e-05,
      "loss": 0.6772,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 523
    },
    {
      "epoch": 0.07125373946151754,
      "grad_norm": 0.5234375,
      "learning_rate": 1.4231292517006802e-05,
      "loss": 0.617,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 524
    },
    {
      "epoch": 0.07138971988033724,
      "grad_norm": 0.48046875,
      "learning_rate": 1.4258503401360545e-05,
      "loss": 0.5855,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 525
    },
    {
      "epoch": 0.07152570029915692,
      "grad_norm": 0.40625,
      "learning_rate": 1.4285714285714287e-05,
      "loss": 0.7098,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 526
    },
    {
      "epoch": 0.07166168071797661,
      "grad_norm": 0.484375,
      "learning_rate": 1.4312925170068029e-05,
      "loss": 0.7419,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 527
    },
    {
      "epoch": 0.0717976611367963,
      "grad_norm": 0.7890625,
      "learning_rate": 1.4340136054421769e-05,
      "loss": 0.7279,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 528
    },
    {
      "epoch": 0.07193364155561599,
      "grad_norm": 0.6953125,
      "learning_rate": 1.436734693877551e-05,
      "loss": 0.5246,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 529
    },
    {
      "epoch": 0.07206962197443569,
      "grad_norm": 0.32421875,
      "learning_rate": 1.4394557823129254e-05,
      "loss": 0.4725,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 530
    },
    {
      "epoch": 0.07220560239325537,
      "grad_norm": 0.5078125,
      "learning_rate": 1.4421768707482994e-05,
      "loss": 0.7705,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 531
    },
    {
      "epoch": 0.07234158281207506,
      "grad_norm": 0.609375,
      "learning_rate": 1.4448979591836736e-05,
      "loss": 0.6609,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 532
    },
    {
      "epoch": 0.07247756323089476,
      "grad_norm": 0.490234375,
      "learning_rate": 1.4476190476190478e-05,
      "loss": 0.7612,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 533
    },
    {
      "epoch": 0.07261354364971444,
      "grad_norm": 0.84765625,
      "learning_rate": 1.4503401360544218e-05,
      "loss": 0.472,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 534
    },
    {
      "epoch": 0.07274952406853413,
      "grad_norm": 0.52734375,
      "learning_rate": 1.4530612244897961e-05,
      "loss": 0.9489,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 535
    },
    {
      "epoch": 0.07288550448735383,
      "grad_norm": 0.388671875,
      "learning_rate": 1.4557823129251703e-05,
      "loss": 0.7168,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 536
    },
    {
      "epoch": 0.07302148490617351,
      "grad_norm": 0.2578125,
      "learning_rate": 1.4585034013605443e-05,
      "loss": 0.3506,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 537
    },
    {
      "epoch": 0.0731574653249932,
      "grad_norm": 0.435546875,
      "learning_rate": 1.4612244897959185e-05,
      "loss": 0.6997,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 538
    },
    {
      "epoch": 0.07329344574381289,
      "grad_norm": 0.45703125,
      "learning_rate": 1.4639455782312926e-05,
      "loss": 0.7103,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 539
    },
    {
      "epoch": 0.07342942616263258,
      "grad_norm": 0.498046875,
      "learning_rate": 1.4666666666666666e-05,
      "loss": 0.6815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 540
    },
    {
      "epoch": 0.07356540658145228,
      "grad_norm": 0.4140625,
      "learning_rate": 1.469387755102041e-05,
      "loss": 0.4292,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 541
    },
    {
      "epoch": 0.07370138700027196,
      "grad_norm": 0.51171875,
      "learning_rate": 1.4721088435374151e-05,
      "loss": 0.8656,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 542
    },
    {
      "epoch": 0.07383736741909165,
      "grad_norm": 0.3828125,
      "learning_rate": 1.4748299319727891e-05,
      "loss": 0.6108,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 543
    },
    {
      "epoch": 0.07397334783791135,
      "grad_norm": 0.349609375,
      "learning_rate": 1.4775510204081633e-05,
      "loss": 0.5869,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 544
    },
    {
      "epoch": 0.07410932825673103,
      "grad_norm": 0.57421875,
      "learning_rate": 1.4802721088435375e-05,
      "loss": 0.7852,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 545
    },
    {
      "epoch": 0.07424530867555072,
      "grad_norm": 0.1982421875,
      "learning_rate": 1.4829931972789118e-05,
      "loss": 0.3281,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 546
    },
    {
      "epoch": 0.07438128909437042,
      "grad_norm": 0.44140625,
      "learning_rate": 1.4857142857142858e-05,
      "loss": 0.7008,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 547
    },
    {
      "epoch": 0.0745172695131901,
      "grad_norm": 0.52734375,
      "learning_rate": 1.48843537414966e-05,
      "loss": 0.7467,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 548
    },
    {
      "epoch": 0.0746532499320098,
      "grad_norm": 1.34375,
      "learning_rate": 1.4911564625850342e-05,
      "loss": 0.9422,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 549
    },
    {
      "epoch": 0.07478923035082947,
      "grad_norm": 0.416015625,
      "learning_rate": 1.4938775510204082e-05,
      "loss": 0.6206,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 550
    },
    {
      "epoch": 0.07492521076964917,
      "grad_norm": 0.6015625,
      "learning_rate": 1.4965986394557825e-05,
      "loss": 0.3831,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 551
    },
    {
      "epoch": 0.07506119118846886,
      "grad_norm": 0.61328125,
      "learning_rate": 1.4993197278911567e-05,
      "loss": 0.5942,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 552
    },
    {
      "epoch": 0.07519717160728855,
      "grad_norm": 0.6953125,
      "learning_rate": 1.5020408163265307e-05,
      "loss": 0.8011,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 553
    },
    {
      "epoch": 0.07533315202610824,
      "grad_norm": 0.5703125,
      "learning_rate": 1.5047619047619049e-05,
      "loss": 0.5545,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 554
    },
    {
      "epoch": 0.07546913244492794,
      "grad_norm": 0.388671875,
      "learning_rate": 1.507482993197279e-05,
      "loss": 0.4624,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 555
    },
    {
      "epoch": 0.07560511286374762,
      "grad_norm": 0.48828125,
      "learning_rate": 1.510204081632653e-05,
      "loss": 0.7686,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 556
    },
    {
      "epoch": 0.07574109328256731,
      "grad_norm": 0.58984375,
      "learning_rate": 1.5129251700680274e-05,
      "loss": 0.7417,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 557
    },
    {
      "epoch": 0.075877073701387,
      "grad_norm": 0.455078125,
      "learning_rate": 1.5156462585034016e-05,
      "loss": 0.9266,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 558
    },
    {
      "epoch": 0.07601305412020669,
      "grad_norm": 0.484375,
      "learning_rate": 1.5183673469387756e-05,
      "loss": 0.7153,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 559
    },
    {
      "epoch": 0.07614903453902638,
      "grad_norm": 0.5234375,
      "learning_rate": 1.5210884353741498e-05,
      "loss": 0.8066,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 560
    },
    {
      "epoch": 0.07628501495784606,
      "grad_norm": 0.60546875,
      "learning_rate": 1.523809523809524e-05,
      "loss": 0.7179,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 561
    },
    {
      "epoch": 0.07642099537666576,
      "grad_norm": 1.5546875,
      "learning_rate": 1.526530612244898e-05,
      "loss": 0.7503,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 562
    },
    {
      "epoch": 0.07655697579548545,
      "grad_norm": 0.498046875,
      "learning_rate": 1.5292517006802724e-05,
      "loss": 0.799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 563
    },
    {
      "epoch": 0.07669295621430514,
      "grad_norm": 0.451171875,
      "learning_rate": 1.5319727891156464e-05,
      "loss": 0.7425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 564
    },
    {
      "epoch": 0.07682893663312483,
      "grad_norm": 0.462890625,
      "learning_rate": 1.5346938775510204e-05,
      "loss": 0.4948,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 565
    },
    {
      "epoch": 0.07696491705194453,
      "grad_norm": 0.326171875,
      "learning_rate": 1.5374149659863945e-05,
      "loss": 0.4196,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 566
    },
    {
      "epoch": 0.07710089747076421,
      "grad_norm": 0.36328125,
      "learning_rate": 1.5401360544217688e-05,
      "loss": 0.6999,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 567
    },
    {
      "epoch": 0.0772368778895839,
      "grad_norm": 0.53125,
      "learning_rate": 1.542857142857143e-05,
      "loss": 0.6751,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 568
    },
    {
      "epoch": 0.07737285830840358,
      "grad_norm": 0.404296875,
      "learning_rate": 1.545578231292517e-05,
      "loss": 0.617,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 569
    },
    {
      "epoch": 0.07750883872722328,
      "grad_norm": 0.359375,
      "learning_rate": 1.548299319727891e-05,
      "loss": 0.7288,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 570
    },
    {
      "epoch": 0.07764481914604297,
      "grad_norm": 0.5859375,
      "learning_rate": 1.5510204081632655e-05,
      "loss": 0.9116,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 571
    },
    {
      "epoch": 0.07778079956486265,
      "grad_norm": 0.67578125,
      "learning_rate": 1.5537414965986395e-05,
      "loss": 0.5809,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 572
    },
    {
      "epoch": 0.07791677998368235,
      "grad_norm": 0.38671875,
      "learning_rate": 1.556462585034014e-05,
      "loss": 0.7199,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 573
    },
    {
      "epoch": 0.07805276040250204,
      "grad_norm": 0.490234375,
      "learning_rate": 1.559183673469388e-05,
      "loss": 0.8745,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 574
    },
    {
      "epoch": 0.07818874082132173,
      "grad_norm": 0.4921875,
      "learning_rate": 1.5619047619047622e-05,
      "loss": 0.7891,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 575
    },
    {
      "epoch": 0.07832472124014142,
      "grad_norm": 0.4609375,
      "learning_rate": 1.5646258503401362e-05,
      "loss": 0.7611,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 576
    },
    {
      "epoch": 0.07846070165896112,
      "grad_norm": 0.412109375,
      "learning_rate": 1.5673469387755102e-05,
      "loss": 0.6354,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 577
    },
    {
      "epoch": 0.0785966820777808,
      "grad_norm": 0.98046875,
      "learning_rate": 1.5700680272108845e-05,
      "loss": 0.7998,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 578
    },
    {
      "epoch": 0.07873266249660049,
      "grad_norm": 0.33984375,
      "learning_rate": 1.5727891156462585e-05,
      "loss": 0.5814,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 579
    },
    {
      "epoch": 0.07886864291542017,
      "grad_norm": 0.40625,
      "learning_rate": 1.575510204081633e-05,
      "loss": 0.8164,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 580
    },
    {
      "epoch": 0.07900462333423987,
      "grad_norm": 0.369140625,
      "learning_rate": 1.578231292517007e-05,
      "loss": 0.7254,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 581
    },
    {
      "epoch": 0.07914060375305956,
      "grad_norm": 0.314453125,
      "learning_rate": 1.580952380952381e-05,
      "loss": 0.5124,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 582
    },
    {
      "epoch": 0.07927658417187924,
      "grad_norm": 0.333984375,
      "learning_rate": 1.5836734693877552e-05,
      "loss": 0.7453,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 583
    },
    {
      "epoch": 0.07941256459069894,
      "grad_norm": 0.45703125,
      "learning_rate": 1.5863945578231296e-05,
      "loss": 0.8633,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 584
    },
    {
      "epoch": 0.07954854500951863,
      "grad_norm": 0.52734375,
      "learning_rate": 1.5891156462585036e-05,
      "loss": 0.5838,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 585
    },
    {
      "epoch": 0.07968452542833832,
      "grad_norm": 0.408203125,
      "learning_rate": 1.5918367346938776e-05,
      "loss": 0.7746,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 586
    },
    {
      "epoch": 0.07982050584715801,
      "grad_norm": 0.625,
      "learning_rate": 1.594557823129252e-05,
      "loss": 0.7969,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 587
    },
    {
      "epoch": 0.0799564862659777,
      "grad_norm": 0.41796875,
      "learning_rate": 1.597278911564626e-05,
      "loss": 0.5858,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 588
    },
    {
      "epoch": 0.08009246668479739,
      "grad_norm": 0.416015625,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.7933,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 589
    },
    {
      "epoch": 0.08022844710361708,
      "grad_norm": 0.50390625,
      "learning_rate": 1.6027210884353743e-05,
      "loss": 0.9272,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 590
    },
    {
      "epoch": 0.08036442752243676,
      "grad_norm": 0.5703125,
      "learning_rate": 1.6054421768707483e-05,
      "loss": 0.8588,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 591
    },
    {
      "epoch": 0.08050040794125646,
      "grad_norm": 0.5234375,
      "learning_rate": 1.6081632653061226e-05,
      "loss": 0.5306,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 592
    },
    {
      "epoch": 0.08063638836007615,
      "grad_norm": 0.37109375,
      "learning_rate": 1.6108843537414966e-05,
      "loss": 0.6553,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 593
    },
    {
      "epoch": 0.08077236877889583,
      "grad_norm": 0.43359375,
      "learning_rate": 1.613605442176871e-05,
      "loss": 0.5894,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 594
    },
    {
      "epoch": 0.08090834919771553,
      "grad_norm": 0.283203125,
      "learning_rate": 1.616326530612245e-05,
      "loss": 0.4754,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 595
    },
    {
      "epoch": 0.08104432961653522,
      "grad_norm": 0.33984375,
      "learning_rate": 1.6190476190476193e-05,
      "loss": 0.7834,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 596
    },
    {
      "epoch": 0.0811803100353549,
      "grad_norm": 0.5625,
      "learning_rate": 1.6217687074829933e-05,
      "loss": 0.5929,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 597
    },
    {
      "epoch": 0.0813162904541746,
      "grad_norm": 0.375,
      "learning_rate": 1.6244897959183673e-05,
      "loss": 0.6294,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 598
    },
    {
      "epoch": 0.0814522708729943,
      "grad_norm": 0.419921875,
      "learning_rate": 1.6272108843537417e-05,
      "loss": 0.6735,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 599
    },
    {
      "epoch": 0.08158825129181398,
      "grad_norm": 0.36328125,
      "learning_rate": 1.629931972789116e-05,
      "loss": 0.6891,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 600
    },
    {
      "epoch": 0.08172423171063367,
      "grad_norm": 0.39453125,
      "learning_rate": 1.63265306122449e-05,
      "loss": 0.721,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 601
    },
    {
      "epoch": 0.08186021212945335,
      "grad_norm": 0.283203125,
      "learning_rate": 1.635374149659864e-05,
      "loss": 0.6235,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 602
    },
    {
      "epoch": 0.08199619254827305,
      "grad_norm": 0.388671875,
      "learning_rate": 1.6380952380952384e-05,
      "loss": 0.7349,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 603
    },
    {
      "epoch": 0.08213217296709274,
      "grad_norm": 0.34375,
      "learning_rate": 1.6408163265306124e-05,
      "loss": 0.5184,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 604
    },
    {
      "epoch": 0.08226815338591242,
      "grad_norm": 0.5546875,
      "learning_rate": 1.6435374149659867e-05,
      "loss": 0.6493,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 605
    },
    {
      "epoch": 0.08240413380473212,
      "grad_norm": 0.4140625,
      "learning_rate": 1.6462585034013607e-05,
      "loss": 0.5878,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 606
    },
    {
      "epoch": 0.08254011422355181,
      "grad_norm": 0.515625,
      "learning_rate": 1.6489795918367347e-05,
      "loss": 0.7202,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 607
    },
    {
      "epoch": 0.0826760946423715,
      "grad_norm": 0.36328125,
      "learning_rate": 1.651700680272109e-05,
      "loss": 0.4784,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 608
    },
    {
      "epoch": 0.08281207506119119,
      "grad_norm": 0.37109375,
      "learning_rate": 1.654421768707483e-05,
      "loss": 0.6064,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 609
    },
    {
      "epoch": 0.08294805548001088,
      "grad_norm": 0.462890625,
      "learning_rate": 1.6571428571428574e-05,
      "loss": 0.7153,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 610
    },
    {
      "epoch": 0.08308403589883057,
      "grad_norm": 0.35546875,
      "learning_rate": 1.6598639455782314e-05,
      "loss": 0.6732,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 611
    },
    {
      "epoch": 0.08322001631765026,
      "grad_norm": 0.462890625,
      "learning_rate": 1.6625850340136057e-05,
      "loss": 0.6934,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 612
    },
    {
      "epoch": 0.08335599673646994,
      "grad_norm": 0.6328125,
      "learning_rate": 1.6653061224489797e-05,
      "loss": 0.8293,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 613
    },
    {
      "epoch": 0.08349197715528964,
      "grad_norm": 0.4375,
      "learning_rate": 1.6680272108843537e-05,
      "loss": 0.6772,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 614
    },
    {
      "epoch": 0.08362795757410933,
      "grad_norm": 0.447265625,
      "learning_rate": 1.670748299319728e-05,
      "loss": 0.6994,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 615
    },
    {
      "epoch": 0.08376393799292901,
      "grad_norm": 0.5,
      "learning_rate": 1.673469387755102e-05,
      "loss": 0.6186,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 616
    },
    {
      "epoch": 0.08389991841174871,
      "grad_norm": 0.6875,
      "learning_rate": 1.6761904761904764e-05,
      "loss": 0.5752,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 617
    },
    {
      "epoch": 0.0840358988305684,
      "grad_norm": 0.451171875,
      "learning_rate": 1.6789115646258504e-05,
      "loss": 0.6973,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 618
    },
    {
      "epoch": 0.08417187924938808,
      "grad_norm": 0.271484375,
      "learning_rate": 1.6816326530612244e-05,
      "loss": 0.5794,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 619
    },
    {
      "epoch": 0.08430785966820778,
      "grad_norm": 0.349609375,
      "learning_rate": 1.6843537414965988e-05,
      "loss": 0.5775,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 620
    },
    {
      "epoch": 0.08444384008702747,
      "grad_norm": 0.431640625,
      "learning_rate": 1.687074829931973e-05,
      "loss": 0.6559,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 621
    },
    {
      "epoch": 0.08457982050584716,
      "grad_norm": 1.046875,
      "learning_rate": 1.689795918367347e-05,
      "loss": 0.4826,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 622
    },
    {
      "epoch": 0.08471580092466685,
      "grad_norm": 0.4140625,
      "learning_rate": 1.692517006802721e-05,
      "loss": 0.6742,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 623
    },
    {
      "epoch": 0.08485178134348653,
      "grad_norm": 0.21484375,
      "learning_rate": 1.6952380952380955e-05,
      "loss": 0.3547,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 624
    },
    {
      "epoch": 0.08498776176230623,
      "grad_norm": 0.73828125,
      "learning_rate": 1.6979591836734695e-05,
      "loss": 0.4362,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 625
    },
    {
      "epoch": 0.08512374218112592,
      "grad_norm": 0.70703125,
      "learning_rate": 1.7006802721088435e-05,
      "loss": 0.8407,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 626
    },
    {
      "epoch": 0.0852597225999456,
      "grad_norm": 0.408203125,
      "learning_rate": 1.7034013605442178e-05,
      "loss": 0.6693,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 627
    },
    {
      "epoch": 0.0853957030187653,
      "grad_norm": 0.396484375,
      "learning_rate": 1.7061224489795922e-05,
      "loss": 0.6473,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 628
    },
    {
      "epoch": 0.085531683437585,
      "grad_norm": 0.4921875,
      "learning_rate": 1.7088435374149662e-05,
      "loss": 0.7824,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 629
    },
    {
      "epoch": 0.08566766385640467,
      "grad_norm": 0.23828125,
      "learning_rate": 1.7115646258503402e-05,
      "loss": 0.5239,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 630
    },
    {
      "epoch": 0.08580364427522437,
      "grad_norm": 0.365234375,
      "learning_rate": 1.7142857142857142e-05,
      "loss": 0.5028,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 631
    },
    {
      "epoch": 0.08593962469404406,
      "grad_norm": 0.5234375,
      "learning_rate": 1.7170068027210885e-05,
      "loss": 0.8488,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 632
    },
    {
      "epoch": 0.08607560511286375,
      "grad_norm": 0.48046875,
      "learning_rate": 1.719727891156463e-05,
      "loss": 0.5894,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 633
    },
    {
      "epoch": 0.08621158553168344,
      "grad_norm": 0.55859375,
      "learning_rate": 1.722448979591837e-05,
      "loss": 0.7474,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 634
    },
    {
      "epoch": 0.08634756595050312,
      "grad_norm": 0.49609375,
      "learning_rate": 1.725170068027211e-05,
      "loss": 0.6912,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 635
    },
    {
      "epoch": 0.08648354636932282,
      "grad_norm": 0.466796875,
      "learning_rate": 1.7278911564625852e-05,
      "loss": 0.6086,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 636
    },
    {
      "epoch": 0.08661952678814251,
      "grad_norm": 0.400390625,
      "learning_rate": 1.7306122448979596e-05,
      "loss": 0.6445,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 637
    },
    {
      "epoch": 0.08675550720696219,
      "grad_norm": 0.71875,
      "learning_rate": 1.7333333333333336e-05,
      "loss": 0.7062,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 638
    },
    {
      "epoch": 0.08689148762578189,
      "grad_norm": 0.453125,
      "learning_rate": 1.7360544217687076e-05,
      "loss": 0.7661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 639
    },
    {
      "epoch": 0.08702746804460158,
      "grad_norm": 0.3984375,
      "learning_rate": 1.738775510204082e-05,
      "loss": 0.5462,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 640
    },
    {
      "epoch": 0.08716344846342126,
      "grad_norm": 0.625,
      "learning_rate": 1.741496598639456e-05,
      "loss": 0.6162,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 641
    },
    {
      "epoch": 0.08729942888224096,
      "grad_norm": 0.53125,
      "learning_rate": 1.74421768707483e-05,
      "loss": 0.5688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 642
    },
    {
      "epoch": 0.08743540930106064,
      "grad_norm": 0.306640625,
      "learning_rate": 1.7469387755102043e-05,
      "loss": 0.4521,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 643
    },
    {
      "epoch": 0.08757138971988034,
      "grad_norm": 0.96875,
      "learning_rate": 1.7496598639455783e-05,
      "loss": 0.8927,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 644
    },
    {
      "epoch": 0.08770737013870003,
      "grad_norm": 0.3828125,
      "learning_rate": 1.7523809523809526e-05,
      "loss": 0.6654,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 645
    },
    {
      "epoch": 0.08784335055751971,
      "grad_norm": 0.458984375,
      "learning_rate": 1.7551020408163266e-05,
      "loss": 0.6686,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 646
    },
    {
      "epoch": 0.0879793309763394,
      "grad_norm": 0.51953125,
      "learning_rate": 1.7578231292517006e-05,
      "loss": 0.6924,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 647
    },
    {
      "epoch": 0.0881153113951591,
      "grad_norm": 0.515625,
      "learning_rate": 1.760544217687075e-05,
      "loss": 0.8356,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 648
    },
    {
      "epoch": 0.08825129181397878,
      "grad_norm": 0.875,
      "learning_rate": 1.7632653061224493e-05,
      "loss": 0.6258,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 649
    },
    {
      "epoch": 0.08838727223279848,
      "grad_norm": 0.484375,
      "learning_rate": 1.7659863945578233e-05,
      "loss": 0.9227,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 650
    },
    {
      "epoch": 0.08852325265161817,
      "grad_norm": 0.390625,
      "learning_rate": 1.7687074829931973e-05,
      "loss": 0.6056,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 651
    },
    {
      "epoch": 0.08865923307043785,
      "grad_norm": 0.5546875,
      "learning_rate": 1.7714285714285717e-05,
      "loss": 0.9338,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 652
    },
    {
      "epoch": 0.08879521348925755,
      "grad_norm": 0.4296875,
      "learning_rate": 1.774149659863946e-05,
      "loss": 0.6589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 653
    },
    {
      "epoch": 0.08893119390807723,
      "grad_norm": 0.609375,
      "learning_rate": 1.77687074829932e-05,
      "loss": 0.8381,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 654
    },
    {
      "epoch": 0.08906717432689693,
      "grad_norm": 0.30859375,
      "learning_rate": 1.779591836734694e-05,
      "loss": 0.6152,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 655
    },
    {
      "epoch": 0.08920315474571662,
      "grad_norm": 0.51171875,
      "learning_rate": 1.782312925170068e-05,
      "loss": 0.8177,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 656
    },
    {
      "epoch": 0.0893391351645363,
      "grad_norm": 0.953125,
      "learning_rate": 1.7850340136054423e-05,
      "loss": 0.7708,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 657
    },
    {
      "epoch": 0.089475115583356,
      "grad_norm": 0.73046875,
      "learning_rate": 1.7877551020408164e-05,
      "loss": 0.7433,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 658
    },
    {
      "epoch": 0.08961109600217569,
      "grad_norm": 1.1640625,
      "learning_rate": 1.7904761904761907e-05,
      "loss": 0.7938,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 659
    },
    {
      "epoch": 0.08974707642099537,
      "grad_norm": 0.416015625,
      "learning_rate": 1.7931972789115647e-05,
      "loss": 0.6092,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 660
    },
    {
      "epoch": 0.08988305683981507,
      "grad_norm": 0.390625,
      "learning_rate": 1.795918367346939e-05,
      "loss": 0.5627,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 661
    },
    {
      "epoch": 0.09001903725863476,
      "grad_norm": 0.38671875,
      "learning_rate": 1.798639455782313e-05,
      "loss": 0.6328,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 662
    },
    {
      "epoch": 0.09015501767745444,
      "grad_norm": 0.400390625,
      "learning_rate": 1.801360544217687e-05,
      "loss": 0.5605,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 663
    },
    {
      "epoch": 0.09029099809627414,
      "grad_norm": 0.302734375,
      "learning_rate": 1.8040816326530614e-05,
      "loss": 0.5127,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 664
    },
    {
      "epoch": 0.09042697851509382,
      "grad_norm": 0.43359375,
      "learning_rate": 1.8068027210884357e-05,
      "loss": 0.6668,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 665
    },
    {
      "epoch": 0.09056295893391351,
      "grad_norm": 0.671875,
      "learning_rate": 1.8095238095238097e-05,
      "loss": 0.8872,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 666
    },
    {
      "epoch": 0.09069893935273321,
      "grad_norm": 0.48828125,
      "learning_rate": 1.8122448979591837e-05,
      "loss": 0.6076,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 667
    },
    {
      "epoch": 0.09083491977155289,
      "grad_norm": 0.515625,
      "learning_rate": 1.8149659863945577e-05,
      "loss": 0.6606,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 668
    },
    {
      "epoch": 0.09097090019037259,
      "grad_norm": 0.462890625,
      "learning_rate": 1.817687074829932e-05,
      "loss": 0.7957,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 669
    },
    {
      "epoch": 0.09110688060919228,
      "grad_norm": 0.4296875,
      "learning_rate": 1.8204081632653064e-05,
      "loss": 0.6429,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 670
    },
    {
      "epoch": 0.09124286102801196,
      "grad_norm": 0.435546875,
      "learning_rate": 1.8231292517006804e-05,
      "loss": 0.512,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 671
    },
    {
      "epoch": 0.09137884144683166,
      "grad_norm": 0.400390625,
      "learning_rate": 1.8258503401360544e-05,
      "loss": 0.8014,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 672
    },
    {
      "epoch": 0.09151482186565135,
      "grad_norm": 0.5078125,
      "learning_rate": 1.8285714285714288e-05,
      "loss": 0.4766,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 673
    },
    {
      "epoch": 0.09165080228447103,
      "grad_norm": 0.396484375,
      "learning_rate": 1.8312925170068028e-05,
      "loss": 0.6714,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 674
    },
    {
      "epoch": 0.09178678270329073,
      "grad_norm": 0.52734375,
      "learning_rate": 1.834013605442177e-05,
      "loss": 0.5931,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 675
    },
    {
      "epoch": 0.09192276312211041,
      "grad_norm": 0.5625,
      "learning_rate": 1.836734693877551e-05,
      "loss": 0.6564,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 676
    },
    {
      "epoch": 0.0920587435409301,
      "grad_norm": 0.4921875,
      "learning_rate": 1.8394557823129255e-05,
      "loss": 0.7808,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 677
    },
    {
      "epoch": 0.0921947239597498,
      "grad_norm": 0.361328125,
      "learning_rate": 1.8421768707482995e-05,
      "loss": 0.6341,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 678
    },
    {
      "epoch": 0.09233070437856948,
      "grad_norm": 0.58203125,
      "learning_rate": 1.8448979591836735e-05,
      "loss": 0.998,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 679
    },
    {
      "epoch": 0.09246668479738918,
      "grad_norm": 0.3515625,
      "learning_rate": 1.8476190476190478e-05,
      "loss": 0.6343,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 680
    },
    {
      "epoch": 0.09260266521620887,
      "grad_norm": 0.470703125,
      "learning_rate": 1.8503401360544218e-05,
      "loss": 0.8618,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 681
    },
    {
      "epoch": 0.09273864563502855,
      "grad_norm": 0.28125,
      "learning_rate": 1.853061224489796e-05,
      "loss": 0.5597,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 682
    },
    {
      "epoch": 0.09287462605384825,
      "grad_norm": 0.435546875,
      "learning_rate": 1.8557823129251702e-05,
      "loss": 0.7568,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 683
    },
    {
      "epoch": 0.09301060647266794,
      "grad_norm": 1.4296875,
      "learning_rate": 1.8585034013605442e-05,
      "loss": 0.7145,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 684
    },
    {
      "epoch": 0.09314658689148762,
      "grad_norm": 0.7421875,
      "learning_rate": 1.8612244897959185e-05,
      "loss": 0.6692,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 685
    },
    {
      "epoch": 0.09328256731030732,
      "grad_norm": 0.53125,
      "learning_rate": 1.863945578231293e-05,
      "loss": 0.6429,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 686
    },
    {
      "epoch": 0.093418547729127,
      "grad_norm": 0.53125,
      "learning_rate": 1.866666666666667e-05,
      "loss": 0.8006,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 687
    },
    {
      "epoch": 0.0935545281479467,
      "grad_norm": 0.396484375,
      "learning_rate": 1.869387755102041e-05,
      "loss": 0.7323,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 688
    },
    {
      "epoch": 0.09369050856676639,
      "grad_norm": 0.37890625,
      "learning_rate": 1.8721088435374152e-05,
      "loss": 0.6424,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 689
    },
    {
      "epoch": 0.09382648898558607,
      "grad_norm": 0.318359375,
      "learning_rate": 1.8748299319727892e-05,
      "loss": 0.5952,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 690
    },
    {
      "epoch": 0.09396246940440577,
      "grad_norm": 0.515625,
      "learning_rate": 1.8775510204081636e-05,
      "loss": 0.7869,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 691
    },
    {
      "epoch": 0.09409844982322546,
      "grad_norm": 0.81640625,
      "learning_rate": 1.8802721088435376e-05,
      "loss": 0.5594,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 692
    },
    {
      "epoch": 0.09423443024204514,
      "grad_norm": 0.40625,
      "learning_rate": 1.8829931972789116e-05,
      "loss": 0.6222,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 693
    },
    {
      "epoch": 0.09437041066086484,
      "grad_norm": 0.283203125,
      "learning_rate": 1.885714285714286e-05,
      "loss": 0.5317,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 694
    },
    {
      "epoch": 0.09450639107968453,
      "grad_norm": 0.28515625,
      "learning_rate": 1.88843537414966e-05,
      "loss": 0.4914,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 695
    },
    {
      "epoch": 0.09464237149850421,
      "grad_norm": 0.353515625,
      "learning_rate": 1.8911564625850343e-05,
      "loss": 0.5547,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 696
    },
    {
      "epoch": 0.09477835191732391,
      "grad_norm": 0.4765625,
      "learning_rate": 1.8938775510204083e-05,
      "loss": 0.7684,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 697
    },
    {
      "epoch": 0.09491433233614359,
      "grad_norm": 0.5234375,
      "learning_rate": 1.8965986394557826e-05,
      "loss": 0.6665,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 698
    },
    {
      "epoch": 0.09505031275496328,
      "grad_norm": 0.5078125,
      "learning_rate": 1.8993197278911566e-05,
      "loss": 0.9193,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 699
    },
    {
      "epoch": 0.09518629317378298,
      "grad_norm": 0.4140625,
      "learning_rate": 1.9020408163265306e-05,
      "loss": 0.5843,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 700
    },
    {
      "epoch": 0.09532227359260266,
      "grad_norm": 0.390625,
      "learning_rate": 1.904761904761905e-05,
      "loss": 0.5347,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 701
    },
    {
      "epoch": 0.09545825401142236,
      "grad_norm": 0.4609375,
      "learning_rate": 1.9074829931972793e-05,
      "loss": 0.7246,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 702
    },
    {
      "epoch": 0.09559423443024205,
      "grad_norm": 0.63671875,
      "learning_rate": 1.9102040816326533e-05,
      "loss": 0.6709,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 703
    },
    {
      "epoch": 0.09573021484906173,
      "grad_norm": 0.2734375,
      "learning_rate": 1.9129251700680273e-05,
      "loss": 0.5265,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 704
    },
    {
      "epoch": 0.09586619526788143,
      "grad_norm": 0.494140625,
      "learning_rate": 1.9156462585034013e-05,
      "loss": 0.7196,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 705
    },
    {
      "epoch": 0.09600217568670112,
      "grad_norm": 1.40625,
      "learning_rate": 1.9183673469387756e-05,
      "loss": 0.734,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 706
    },
    {
      "epoch": 0.0961381561055208,
      "grad_norm": 0.6640625,
      "learning_rate": 1.92108843537415e-05,
      "loss": 0.7389,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 707
    },
    {
      "epoch": 0.0962741365243405,
      "grad_norm": 0.3125,
      "learning_rate": 1.923809523809524e-05,
      "loss": 0.5981,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 708
    },
    {
      "epoch": 0.09641011694316018,
      "grad_norm": 0.5234375,
      "learning_rate": 1.926530612244898e-05,
      "loss": 0.6906,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 709
    },
    {
      "epoch": 0.09654609736197987,
      "grad_norm": 0.341796875,
      "learning_rate": 1.9292517006802723e-05,
      "loss": 0.6302,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 710
    },
    {
      "epoch": 0.09668207778079957,
      "grad_norm": 0.33203125,
      "learning_rate": 1.9319727891156463e-05,
      "loss": 0.6865,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 711
    },
    {
      "epoch": 0.09681805819961925,
      "grad_norm": 0.330078125,
      "learning_rate": 1.9346938775510207e-05,
      "loss": 0.5322,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 712
    },
    {
      "epoch": 0.09695403861843895,
      "grad_norm": 0.39453125,
      "learning_rate": 1.9374149659863947e-05,
      "loss": 0.7124,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 713
    },
    {
      "epoch": 0.09709001903725864,
      "grad_norm": 0.921875,
      "learning_rate": 1.940136054421769e-05,
      "loss": 0.8219,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 714
    },
    {
      "epoch": 0.09722599945607832,
      "grad_norm": 0.27734375,
      "learning_rate": 1.942857142857143e-05,
      "loss": 0.45,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 715
    },
    {
      "epoch": 0.09736197987489802,
      "grad_norm": 0.4375,
      "learning_rate": 1.945578231292517e-05,
      "loss": 0.6948,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 716
    },
    {
      "epoch": 0.09749796029371771,
      "grad_norm": 0.3671875,
      "learning_rate": 1.9482993197278914e-05,
      "loss": 0.6908,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 717
    },
    {
      "epoch": 0.09763394071253739,
      "grad_norm": 0.8359375,
      "learning_rate": 1.9510204081632654e-05,
      "loss": 0.638,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 718
    },
    {
      "epoch": 0.09776992113135709,
      "grad_norm": 0.33203125,
      "learning_rate": 1.9537414965986397e-05,
      "loss": 0.5465,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 719
    },
    {
      "epoch": 0.09790590155017677,
      "grad_norm": 0.3125,
      "learning_rate": 1.9564625850340137e-05,
      "loss": 0.5534,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 720
    },
    {
      "epoch": 0.09804188196899646,
      "grad_norm": 0.421875,
      "learning_rate": 1.9591836734693877e-05,
      "loss": 0.7804,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 721
    },
    {
      "epoch": 0.09817786238781616,
      "grad_norm": 0.6953125,
      "learning_rate": 1.961904761904762e-05,
      "loss": 0.6732,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 722
    },
    {
      "epoch": 0.09831384280663584,
      "grad_norm": 0.427734375,
      "learning_rate": 1.9646258503401364e-05,
      "loss": 0.7542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 723
    },
    {
      "epoch": 0.09844982322545553,
      "grad_norm": 0.306640625,
      "learning_rate": 1.9673469387755104e-05,
      "loss": 0.4567,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 724
    },
    {
      "epoch": 0.09858580364427523,
      "grad_norm": 0.392578125,
      "learning_rate": 1.9700680272108844e-05,
      "loss": 0.5392,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 725
    },
    {
      "epoch": 0.09872178406309491,
      "grad_norm": 0.359375,
      "learning_rate": 1.9727891156462588e-05,
      "loss": 0.597,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 726
    },
    {
      "epoch": 0.0988577644819146,
      "grad_norm": 0.40234375,
      "learning_rate": 1.9755102040816328e-05,
      "loss": 0.5068,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 727
    },
    {
      "epoch": 0.09899374490073429,
      "grad_norm": 0.376953125,
      "learning_rate": 1.978231292517007e-05,
      "loss": 0.4753,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 728
    },
    {
      "epoch": 0.09912972531955398,
      "grad_norm": 0.333984375,
      "learning_rate": 1.980952380952381e-05,
      "loss": 0.6844,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 729
    },
    {
      "epoch": 0.09926570573837368,
      "grad_norm": 0.4296875,
      "learning_rate": 1.983673469387755e-05,
      "loss": 0.7507,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 730
    },
    {
      "epoch": 0.09940168615719336,
      "grad_norm": 0.44921875,
      "learning_rate": 1.9863945578231295e-05,
      "loss": 0.813,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 731
    },
    {
      "epoch": 0.09953766657601305,
      "grad_norm": 0.265625,
      "learning_rate": 1.9891156462585035e-05,
      "loss": 0.4027,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 732
    },
    {
      "epoch": 0.09967364699483275,
      "grad_norm": 0.275390625,
      "learning_rate": 1.9918367346938775e-05,
      "loss": 0.5864,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 733
    },
    {
      "epoch": 0.09980962741365243,
      "grad_norm": 0.5078125,
      "learning_rate": 1.9945578231292518e-05,
      "loss": 0.9105,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 734
    },
    {
      "epoch": 0.09994560783247212,
      "grad_norm": 0.498046875,
      "learning_rate": 1.997278911564626e-05,
      "loss": 0.7962,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 735
    },
    {
      "epoch": 0.10008158825129182,
      "grad_norm": 0.427734375,
      "learning_rate": 2e-05,
      "loss": 0.6499,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 736
    },
    {
      "epoch": 0.1002175686701115,
      "grad_norm": 0.62109375,
      "learning_rate": 1.9999998872938873e-05,
      "loss": 0.6597,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 737
    },
    {
      "epoch": 0.1003535490889312,
      "grad_norm": 0.34765625,
      "learning_rate": 1.9999995491755736e-05,
      "loss": 0.5028,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 738
    },
    {
      "epoch": 0.10048952950775088,
      "grad_norm": 0.396484375,
      "learning_rate": 1.9999989856451352e-05,
      "loss": 0.8711,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 739
    },
    {
      "epoch": 0.10062550992657057,
      "grad_norm": 0.33984375,
      "learning_rate": 1.9999981967027e-05,
      "loss": 0.592,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 740
    },
    {
      "epoch": 0.10076149034539027,
      "grad_norm": 0.341796875,
      "learning_rate": 1.9999971823484444e-05,
      "loss": 0.5791,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 741
    },
    {
      "epoch": 0.10089747076420995,
      "grad_norm": 0.3203125,
      "learning_rate": 1.9999959425825983e-05,
      "loss": 0.6021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 742
    },
    {
      "epoch": 0.10103345118302964,
      "grad_norm": 0.5859375,
      "learning_rate": 1.9999944774054406e-05,
      "loss": 0.9653,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 743
    },
    {
      "epoch": 0.10116943160184934,
      "grad_norm": 1.53125,
      "learning_rate": 1.9999927868173018e-05,
      "loss": 0.8663,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 744
    },
    {
      "epoch": 0.10130541202066902,
      "grad_norm": 0.267578125,
      "learning_rate": 1.9999908708185626e-05,
      "loss": 0.4526,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 745
    },
    {
      "epoch": 0.10144139243948871,
      "grad_norm": 0.6171875,
      "learning_rate": 1.9999887294096557e-05,
      "loss": 0.569,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 746
    },
    {
      "epoch": 0.10157737285830841,
      "grad_norm": 0.283203125,
      "learning_rate": 1.999986362591063e-05,
      "loss": 0.4289,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 747
    },
    {
      "epoch": 0.10171335327712809,
      "grad_norm": 0.412109375,
      "learning_rate": 1.999983770363318e-05,
      "loss": 0.6408,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 748
    },
    {
      "epoch": 0.10184933369594779,
      "grad_norm": 0.3203125,
      "learning_rate": 1.9999809527270053e-05,
      "loss": 0.5498,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 749
    },
    {
      "epoch": 0.10198531411476747,
      "grad_norm": 0.458984375,
      "learning_rate": 1.9999779096827603e-05,
      "loss": 0.7254,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 750
    },
    {
      "epoch": 0.10212129453358716,
      "grad_norm": 0.515625,
      "learning_rate": 1.9999746412312684e-05,
      "loss": 0.8105,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 751
    },
    {
      "epoch": 0.10225727495240686,
      "grad_norm": 0.578125,
      "learning_rate": 1.9999711473732667e-05,
      "loss": 0.7323,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 752
    },
    {
      "epoch": 0.10239325537122654,
      "grad_norm": 0.6640625,
      "learning_rate": 1.9999674281095425e-05,
      "loss": 0.8099,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 753
    },
    {
      "epoch": 0.10252923579004623,
      "grad_norm": 0.318359375,
      "learning_rate": 1.9999634834409348e-05,
      "loss": 0.5698,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 754
    },
    {
      "epoch": 0.10266521620886593,
      "grad_norm": 0.515625,
      "learning_rate": 1.999959313368332e-05,
      "loss": 0.5747,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 755
    },
    {
      "epoch": 0.10280119662768561,
      "grad_norm": 0.59375,
      "learning_rate": 1.9999549178926743e-05,
      "loss": 0.7821,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 756
    },
    {
      "epoch": 0.1029371770465053,
      "grad_norm": 0.44140625,
      "learning_rate": 1.9999502970149524e-05,
      "loss": 0.5861,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 757
    },
    {
      "epoch": 0.103073157465325,
      "grad_norm": 0.408203125,
      "learning_rate": 1.9999454507362085e-05,
      "loss": 0.7007,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 758
    },
    {
      "epoch": 0.10320913788414468,
      "grad_norm": 0.34765625,
      "learning_rate": 1.9999403790575344e-05,
      "loss": 0.5633,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 759
    },
    {
      "epoch": 0.10334511830296438,
      "grad_norm": 0.45703125,
      "learning_rate": 1.9999350819800733e-05,
      "loss": 0.7147,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 760
    },
    {
      "epoch": 0.10348109872178406,
      "grad_norm": 0.34765625,
      "learning_rate": 1.99992955950502e-05,
      "loss": 0.4688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 761
    },
    {
      "epoch": 0.10361707914060375,
      "grad_norm": 0.96484375,
      "learning_rate": 1.999923811633618e-05,
      "loss": 0.8335,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 762
    },
    {
      "epoch": 0.10375305955942345,
      "grad_norm": 0.515625,
      "learning_rate": 1.9999178383671644e-05,
      "loss": 0.582,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 763
    },
    {
      "epoch": 0.10388903997824313,
      "grad_norm": 0.41015625,
      "learning_rate": 1.9999116397070045e-05,
      "loss": 0.5915,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 764
    },
    {
      "epoch": 0.10402502039706282,
      "grad_norm": 0.76953125,
      "learning_rate": 1.9999052156545365e-05,
      "loss": 0.8001,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 765
    },
    {
      "epoch": 0.10416100081588252,
      "grad_norm": 0.39453125,
      "learning_rate": 1.9998985662112077e-05,
      "loss": 0.6354,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 766
    },
    {
      "epoch": 0.1042969812347022,
      "grad_norm": 0.392578125,
      "learning_rate": 1.999891691378517e-05,
      "loss": 0.6974,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 767
    },
    {
      "epoch": 0.1044329616535219,
      "grad_norm": 0.328125,
      "learning_rate": 1.9998845911580143e-05,
      "loss": 0.4718,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 768
    },
    {
      "epoch": 0.10456894207234159,
      "grad_norm": 0.55078125,
      "learning_rate": 1.9998772655513007e-05,
      "loss": 0.8677,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 769
    },
    {
      "epoch": 0.10470492249116127,
      "grad_norm": 0.65625,
      "learning_rate": 1.999869714560026e-05,
      "loss": 0.8179,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 770
    },
    {
      "epoch": 0.10484090290998097,
      "grad_norm": 0.77734375,
      "learning_rate": 1.999861938185894e-05,
      "loss": 0.5334,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 771
    },
    {
      "epoch": 0.10497688332880065,
      "grad_norm": 0.3359375,
      "learning_rate": 1.999853936430656e-05,
      "loss": 0.6322,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 772
    },
    {
      "epoch": 0.10511286374762034,
      "grad_norm": 0.56640625,
      "learning_rate": 1.9998457092961172e-05,
      "loss": 0.7773,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 773
    },
    {
      "epoch": 0.10524884416644004,
      "grad_norm": 0.484375,
      "learning_rate": 1.9998372567841308e-05,
      "loss": 0.772,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 774
    },
    {
      "epoch": 0.10538482458525972,
      "grad_norm": 0.400390625,
      "learning_rate": 1.9998285788966027e-05,
      "loss": 0.6842,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 775
    },
    {
      "epoch": 0.10552080500407941,
      "grad_norm": 0.326171875,
      "learning_rate": 1.9998196756354892e-05,
      "loss": 0.6032,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 776
    },
    {
      "epoch": 0.10565678542289911,
      "grad_norm": 0.359375,
      "learning_rate": 1.9998105470027965e-05,
      "loss": 0.6649,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 777
    },
    {
      "epoch": 0.10579276584171879,
      "grad_norm": 0.2255859375,
      "learning_rate": 1.999801193000583e-05,
      "loss": 0.3735,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 778
    },
    {
      "epoch": 0.10592874626053848,
      "grad_norm": 0.51953125,
      "learning_rate": 1.999791613630957e-05,
      "loss": 0.6261,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 779
    },
    {
      "epoch": 0.10606472667935818,
      "grad_norm": 0.53515625,
      "learning_rate": 1.9997818088960777e-05,
      "loss": 0.7533,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 780
    },
    {
      "epoch": 0.10620070709817786,
      "grad_norm": 0.609375,
      "learning_rate": 1.999771778798155e-05,
      "loss": 0.6027,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 781
    },
    {
      "epoch": 0.10633668751699756,
      "grad_norm": 0.3046875,
      "learning_rate": 1.9997615233394505e-05,
      "loss": 0.4487,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 782
    },
    {
      "epoch": 0.10647266793581724,
      "grad_norm": 0.35546875,
      "learning_rate": 1.9997510425222754e-05,
      "loss": 0.5986,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 783
    },
    {
      "epoch": 0.10660864835463693,
      "grad_norm": 0.78515625,
      "learning_rate": 1.999740336348992e-05,
      "loss": 0.6808,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 784
    },
    {
      "epoch": 0.10674462877345663,
      "grad_norm": 2.78125,
      "learning_rate": 1.9997294048220142e-05,
      "loss": 0.8021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 785
    },
    {
      "epoch": 0.10688060919227631,
      "grad_norm": 0.466796875,
      "learning_rate": 1.9997182479438056e-05,
      "loss": 0.7575,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 786
    },
    {
      "epoch": 0.107016589611096,
      "grad_norm": 0.5,
      "learning_rate": 1.9997068657168812e-05,
      "loss": 0.6589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 787
    },
    {
      "epoch": 0.1071525700299157,
      "grad_norm": 0.474609375,
      "learning_rate": 1.999695258143807e-05,
      "loss": 0.7007,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 788
    },
    {
      "epoch": 0.10728855044873538,
      "grad_norm": 0.46484375,
      "learning_rate": 1.999683425227199e-05,
      "loss": 0.8709,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 789
    },
    {
      "epoch": 0.10742453086755507,
      "grad_norm": 0.369140625,
      "learning_rate": 1.999671366969725e-05,
      "loss": 0.8291,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 790
    },
    {
      "epoch": 0.10756051128637477,
      "grad_norm": 0.302734375,
      "learning_rate": 1.999659083374103e-05,
      "loss": 0.6457,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 791
    },
    {
      "epoch": 0.10769649170519445,
      "grad_norm": 0.361328125,
      "learning_rate": 1.999646574443101e-05,
      "loss": 0.6382,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 792
    },
    {
      "epoch": 0.10783247212401414,
      "grad_norm": 0.66796875,
      "learning_rate": 1.9996338401795395e-05,
      "loss": 0.8293,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 793
    },
    {
      "epoch": 0.10796845254283383,
      "grad_norm": 0.447265625,
      "learning_rate": 1.9996208805862893e-05,
      "loss": 0.7599,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 794
    },
    {
      "epoch": 0.10810443296165352,
      "grad_norm": 0.44140625,
      "learning_rate": 1.999607695666271e-05,
      "loss": 0.8091,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 795
    },
    {
      "epoch": 0.10824041338047322,
      "grad_norm": 0.39453125,
      "learning_rate": 1.9995942854224566e-05,
      "loss": 0.4757,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 796
    },
    {
      "epoch": 0.1083763937992929,
      "grad_norm": 0.365234375,
      "learning_rate": 1.9995806498578693e-05,
      "loss": 0.5868,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 797
    },
    {
      "epoch": 0.10851237421811259,
      "grad_norm": 0.279296875,
      "learning_rate": 1.9995667889755828e-05,
      "loss": 0.5632,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 798
    },
    {
      "epoch": 0.10864835463693229,
      "grad_norm": 0.49609375,
      "learning_rate": 1.9995527027787213e-05,
      "loss": 0.7951,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 799
    },
    {
      "epoch": 0.10878433505575197,
      "grad_norm": 0.337890625,
      "learning_rate": 1.9995383912704596e-05,
      "loss": 0.6177,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 800
    },
    {
      "epoch": 0.10892031547457166,
      "grad_norm": 0.5390625,
      "learning_rate": 1.999523854454024e-05,
      "loss": 0.8128,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 801
    },
    {
      "epoch": 0.10905629589339136,
      "grad_norm": 0.41015625,
      "learning_rate": 1.9995090923326916e-05,
      "loss": 0.6994,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 802
    },
    {
      "epoch": 0.10919227631221104,
      "grad_norm": 0.3984375,
      "learning_rate": 1.9994941049097898e-05,
      "loss": 0.6922,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 803
    },
    {
      "epoch": 0.10932825673103073,
      "grad_norm": 0.57421875,
      "learning_rate": 1.9994788921886968e-05,
      "loss": 0.5726,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 804
    },
    {
      "epoch": 0.10946423714985042,
      "grad_norm": 0.53515625,
      "learning_rate": 1.999463454172842e-05,
      "loss": 0.6491,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 805
    },
    {
      "epoch": 0.10960021756867011,
      "grad_norm": 0.3125,
      "learning_rate": 1.9994477908657044e-05,
      "loss": 0.6577,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 806
    },
    {
      "epoch": 0.1097361979874898,
      "grad_norm": 0.419921875,
      "learning_rate": 1.999431902270816e-05,
      "loss": 0.7557,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 807
    },
    {
      "epoch": 0.10987217840630949,
      "grad_norm": 0.4765625,
      "learning_rate": 1.9994157883917574e-05,
      "loss": 0.5723,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 808
    },
    {
      "epoch": 0.11000815882512918,
      "grad_norm": 0.451171875,
      "learning_rate": 1.9993994492321613e-05,
      "loss": 0.6719,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 809
    },
    {
      "epoch": 0.11014413924394888,
      "grad_norm": 0.54296875,
      "learning_rate": 1.9993828847957106e-05,
      "loss": 0.7189,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 810
    },
    {
      "epoch": 0.11028011966276856,
      "grad_norm": 0.451171875,
      "learning_rate": 1.999366095086139e-05,
      "loss": 0.5809,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 811
    },
    {
      "epoch": 0.11041610008158825,
      "grad_norm": 0.65234375,
      "learning_rate": 1.9993490801072316e-05,
      "loss": 0.6644,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 812
    },
    {
      "epoch": 0.11055208050040793,
      "grad_norm": 0.53515625,
      "learning_rate": 1.9993318398628233e-05,
      "loss": 0.673,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 813
    },
    {
      "epoch": 0.11068806091922763,
      "grad_norm": 0.44921875,
      "learning_rate": 1.9993143743568e-05,
      "loss": 0.7445,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 814
    },
    {
      "epoch": 0.11082404133804732,
      "grad_norm": 0.494140625,
      "learning_rate": 1.9992966835930995e-05,
      "loss": 0.8888,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 815
    },
    {
      "epoch": 0.110960021756867,
      "grad_norm": 0.349609375,
      "learning_rate": 1.999278767575709e-05,
      "loss": 0.5911,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 816
    },
    {
      "epoch": 0.1110960021756867,
      "grad_norm": 0.52734375,
      "learning_rate": 1.9992606263086664e-05,
      "loss": 0.5924,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 817
    },
    {
      "epoch": 0.1112319825945064,
      "grad_norm": 0.66796875,
      "learning_rate": 1.9992422597960624e-05,
      "loss": 0.6971,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 818
    },
    {
      "epoch": 0.11136796301332608,
      "grad_norm": 0.66796875,
      "learning_rate": 1.9992236680420357e-05,
      "loss": 0.7976,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 819
    },
    {
      "epoch": 0.11150394343214577,
      "grad_norm": 0.7109375,
      "learning_rate": 1.9992048510507776e-05,
      "loss": 0.8423,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 820
    },
    {
      "epoch": 0.11163992385096547,
      "grad_norm": 0.58984375,
      "learning_rate": 1.99918580882653e-05,
      "loss": 0.9198,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 821
    },
    {
      "epoch": 0.11177590426978515,
      "grad_norm": 0.47265625,
      "learning_rate": 1.999166541373585e-05,
      "loss": 0.7736,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 822
    },
    {
      "epoch": 0.11191188468860484,
      "grad_norm": 0.65625,
      "learning_rate": 1.9991470486962858e-05,
      "loss": 0.8368,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 823
    },
    {
      "epoch": 0.11204786510742452,
      "grad_norm": 0.28515625,
      "learning_rate": 1.999127330799026e-05,
      "loss": 0.4556,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 824
    },
    {
      "epoch": 0.11218384552624422,
      "grad_norm": 0.46484375,
      "learning_rate": 1.9991073876862502e-05,
      "loss": 0.6619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 825
    },
    {
      "epoch": 0.11231982594506391,
      "grad_norm": 0.42578125,
      "learning_rate": 1.9990872193624544e-05,
      "loss": 0.5531,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 826
    },
    {
      "epoch": 0.1124558063638836,
      "grad_norm": 0.39453125,
      "learning_rate": 1.999066825832184e-05,
      "loss": 0.5579,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 827
    },
    {
      "epoch": 0.11259178678270329,
      "grad_norm": 0.6953125,
      "learning_rate": 1.9990462071000366e-05,
      "loss": 0.5505,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 828
    },
    {
      "epoch": 0.11272776720152299,
      "grad_norm": 0.59375,
      "learning_rate": 1.99902536317066e-05,
      "loss": 0.884,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 829
    },
    {
      "epoch": 0.11286374762034267,
      "grad_norm": 0.365234375,
      "learning_rate": 1.999004294048752e-05,
      "loss": 0.679,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 830
    },
    {
      "epoch": 0.11299972803916236,
      "grad_norm": 0.384765625,
      "learning_rate": 1.9989829997390625e-05,
      "loss": 0.7008,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 831
    },
    {
      "epoch": 0.11313570845798206,
      "grad_norm": 0.431640625,
      "learning_rate": 1.998961480246391e-05,
      "loss": 0.5817,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 832
    },
    {
      "epoch": 0.11327168887680174,
      "grad_norm": 0.54296875,
      "learning_rate": 1.9989397355755885e-05,
      "loss": 0.8073,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 833
    },
    {
      "epoch": 0.11340766929562143,
      "grad_norm": 0.578125,
      "learning_rate": 1.9989177657315567e-05,
      "loss": 0.6942,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 834
    },
    {
      "epoch": 0.11354364971444111,
      "grad_norm": 0.435546875,
      "learning_rate": 1.998895570719247e-05,
      "loss": 0.8542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 835
    },
    {
      "epoch": 0.11367963013326081,
      "grad_norm": 0.578125,
      "learning_rate": 1.998873150543664e-05,
      "loss": 0.461,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 836
    },
    {
      "epoch": 0.1138156105520805,
      "grad_norm": 0.275390625,
      "learning_rate": 1.9988505052098598e-05,
      "loss": 0.4388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 837
    },
    {
      "epoch": 0.11395159097090019,
      "grad_norm": 0.404296875,
      "learning_rate": 1.9988276347229404e-05,
      "loss": 0.6938,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 838
    },
    {
      "epoch": 0.11408757138971988,
      "grad_norm": 0.50390625,
      "learning_rate": 1.9988045390880598e-05,
      "loss": 0.63,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 839
    },
    {
      "epoch": 0.11422355180853958,
      "grad_norm": 0.3125,
      "learning_rate": 1.998781218310425e-05,
      "loss": 0.5101,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 840
    },
    {
      "epoch": 0.11435953222735926,
      "grad_norm": 0.5859375,
      "learning_rate": 1.9987576723952923e-05,
      "loss": 0.6717,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 841
    },
    {
      "epoch": 0.11449551264617895,
      "grad_norm": 0.65234375,
      "learning_rate": 1.9987339013479693e-05,
      "loss": 0.5623,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 842
    },
    {
      "epoch": 0.11463149306499865,
      "grad_norm": 0.427734375,
      "learning_rate": 1.9987099051738142e-05,
      "loss": 0.7295,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 843
    },
    {
      "epoch": 0.11476747348381833,
      "grad_norm": 0.65234375,
      "learning_rate": 1.998685683878236e-05,
      "loss": 0.5285,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 844
    },
    {
      "epoch": 0.11490345390263802,
      "grad_norm": 0.3984375,
      "learning_rate": 1.998661237466695e-05,
      "loss": 0.6227,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 845
    },
    {
      "epoch": 0.1150394343214577,
      "grad_norm": 0.439453125,
      "learning_rate": 1.9986365659447015e-05,
      "loss": 0.714,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 846
    },
    {
      "epoch": 0.1151754147402774,
      "grad_norm": 0.5390625,
      "learning_rate": 1.9986116693178163e-05,
      "loss": 0.6069,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 847
    },
    {
      "epoch": 0.1153113951590971,
      "grad_norm": 0.3671875,
      "learning_rate": 1.9985865475916522e-05,
      "loss": 0.7363,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 848
    },
    {
      "epoch": 0.11544737557791677,
      "grad_norm": 2.609375,
      "learning_rate": 1.998561200771871e-05,
      "loss": 0.9087,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 849
    },
    {
      "epoch": 0.11558335599673647,
      "grad_norm": 0.4765625,
      "learning_rate": 1.9985356288641865e-05,
      "loss": 0.7489,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 850
    },
    {
      "epoch": 0.11571933641555616,
      "grad_norm": 0.62109375,
      "learning_rate": 1.9985098318743635e-05,
      "loss": 0.9269,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 851
    },
    {
      "epoch": 0.11585531683437585,
      "grad_norm": 0.318359375,
      "learning_rate": 1.9984838098082166e-05,
      "loss": 0.491,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 852
    },
    {
      "epoch": 0.11599129725319554,
      "grad_norm": 0.431640625,
      "learning_rate": 1.9984575626716113e-05,
      "loss": 0.6169,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 853
    },
    {
      "epoch": 0.11612727767201524,
      "grad_norm": 0.36328125,
      "learning_rate": 1.998431090470464e-05,
      "loss": 0.6247,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 854
    },
    {
      "epoch": 0.11626325809083492,
      "grad_norm": 0.384765625,
      "learning_rate": 1.9984043932107423e-05,
      "loss": 0.6287,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 855
    },
    {
      "epoch": 0.11639923850965461,
      "grad_norm": 0.515625,
      "learning_rate": 1.9983774708984634e-05,
      "loss": 0.7202,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 856
    },
    {
      "epoch": 0.1165352189284743,
      "grad_norm": 0.37890625,
      "learning_rate": 1.9983503235396962e-05,
      "loss": 0.5926,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 857
    },
    {
      "epoch": 0.11667119934729399,
      "grad_norm": 0.44921875,
      "learning_rate": 1.9983229511405606e-05,
      "loss": 0.6304,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 858
    },
    {
      "epoch": 0.11680717976611368,
      "grad_norm": 0.35546875,
      "learning_rate": 1.998295353707226e-05,
      "loss": 0.647,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 859
    },
    {
      "epoch": 0.11694316018493336,
      "grad_norm": 0.8828125,
      "learning_rate": 1.998267531245914e-05,
      "loss": 0.7101,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 860
    },
    {
      "epoch": 0.11707914060375306,
      "grad_norm": 0.60546875,
      "learning_rate": 1.998239483762895e-05,
      "loss": 0.4635,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 861
    },
    {
      "epoch": 0.11721512102257275,
      "grad_norm": 0.41015625,
      "learning_rate": 1.998211211264492e-05,
      "loss": 0.6068,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 862
    },
    {
      "epoch": 0.11735110144139244,
      "grad_norm": 0.80859375,
      "learning_rate": 1.9981827137570774e-05,
      "loss": 0.6992,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 863
    },
    {
      "epoch": 0.11748708186021213,
      "grad_norm": 0.30859375,
      "learning_rate": 1.9981539912470754e-05,
      "loss": 0.5244,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 864
    },
    {
      "epoch": 0.11762306227903183,
      "grad_norm": 0.357421875,
      "learning_rate": 1.9981250437409607e-05,
      "loss": 0.526,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 865
    },
    {
      "epoch": 0.11775904269785151,
      "grad_norm": 0.328125,
      "learning_rate": 1.9980958712452577e-05,
      "loss": 0.584,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 866
    },
    {
      "epoch": 0.1178950231166712,
      "grad_norm": 0.337890625,
      "learning_rate": 1.9980664737665423e-05,
      "loss": 0.6514,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 867
    },
    {
      "epoch": 0.11803100353549088,
      "grad_norm": 0.291015625,
      "learning_rate": 1.9980368513114418e-05,
      "loss": 0.5661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 868
    },
    {
      "epoch": 0.11816698395431058,
      "grad_norm": 0.6015625,
      "learning_rate": 1.9980070038866324e-05,
      "loss": 0.8156,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 869
    },
    {
      "epoch": 0.11830296437313027,
      "grad_norm": 0.54296875,
      "learning_rate": 1.9979769314988432e-05,
      "loss": 0.8319,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 870
    },
    {
      "epoch": 0.11843894479194995,
      "grad_norm": 0.412109375,
      "learning_rate": 1.9979466341548517e-05,
      "loss": 0.528,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 871
    },
    {
      "epoch": 0.11857492521076965,
      "grad_norm": 0.400390625,
      "learning_rate": 1.997916111861488e-05,
      "loss": 0.8024,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 872
    },
    {
      "epoch": 0.11871090562958934,
      "grad_norm": 0.427734375,
      "learning_rate": 1.997885364625632e-05,
      "loss": 0.7461,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 873
    },
    {
      "epoch": 0.11884688604840903,
      "grad_norm": 0.404296875,
      "learning_rate": 1.997854392454215e-05,
      "loss": 0.6426,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 874
    },
    {
      "epoch": 0.11898286646722872,
      "grad_norm": 0.390625,
      "learning_rate": 1.997823195354218e-05,
      "loss": 0.7282,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 875
    },
    {
      "epoch": 0.11911884688604842,
      "grad_norm": 1.4921875,
      "learning_rate": 1.997791773332673e-05,
      "loss": 0.8179,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 876
    },
    {
      "epoch": 0.1192548273048681,
      "grad_norm": 0.35546875,
      "learning_rate": 1.9977601263966637e-05,
      "loss": 0.7039,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 877
    },
    {
      "epoch": 0.11939080772368779,
      "grad_norm": 0.6171875,
      "learning_rate": 1.9977282545533227e-05,
      "loss": 0.8731,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 878
    },
    {
      "epoch": 0.11952678814250747,
      "grad_norm": 0.3125,
      "learning_rate": 1.9976961578098352e-05,
      "loss": 0.5633,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 879
    },
    {
      "epoch": 0.11966276856132717,
      "grad_norm": 0.6328125,
      "learning_rate": 1.9976638361734356e-05,
      "loss": 0.7718,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 880
    },
    {
      "epoch": 0.11979874898014686,
      "grad_norm": 0.7578125,
      "learning_rate": 1.9976312896514098e-05,
      "loss": 0.8467,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 881
    },
    {
      "epoch": 0.11993472939896654,
      "grad_norm": 0.6640625,
      "learning_rate": 1.9975985182510943e-05,
      "loss": 0.7064,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 882
    },
    {
      "epoch": 0.12007070981778624,
      "grad_norm": 0.375,
      "learning_rate": 1.9975655219798757e-05,
      "loss": 0.5734,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 883
    },
    {
      "epoch": 0.12020669023660593,
      "grad_norm": 0.275390625,
      "learning_rate": 1.9975323008451926e-05,
      "loss": 0.4199,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 884
    },
    {
      "epoch": 0.12034267065542562,
      "grad_norm": 0.69921875,
      "learning_rate": 1.9974988548545328e-05,
      "loss": 0.6841,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 885
    },
    {
      "epoch": 0.12047865107424531,
      "grad_norm": 0.4609375,
      "learning_rate": 1.9974651840154355e-05,
      "loss": 0.7264,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 886
    },
    {
      "epoch": 0.120614631493065,
      "grad_norm": 0.546875,
      "learning_rate": 1.9974312883354908e-05,
      "loss": 0.8101,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 887
    },
    {
      "epoch": 0.12075061191188469,
      "grad_norm": 0.337890625,
      "learning_rate": 1.997397167822339e-05,
      "loss": 0.5518,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 888
    },
    {
      "epoch": 0.12088659233070438,
      "grad_norm": 0.443359375,
      "learning_rate": 1.9973628224836715e-05,
      "loss": 0.6507,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 889
    },
    {
      "epoch": 0.12102257274952406,
      "grad_norm": 0.474609375,
      "learning_rate": 1.9973282523272297e-05,
      "loss": 0.6854,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 890
    },
    {
      "epoch": 0.12115855316834376,
      "grad_norm": 0.48828125,
      "learning_rate": 1.997293457360807e-05,
      "loss": 0.8841,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 891
    },
    {
      "epoch": 0.12129453358716345,
      "grad_norm": 0.5078125,
      "learning_rate": 1.9972584375922453e-05,
      "loss": 0.8254,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 892
    },
    {
      "epoch": 0.12143051400598313,
      "grad_norm": 0.447265625,
      "learning_rate": 1.9972231930294392e-05,
      "loss": 0.7121,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 893
    },
    {
      "epoch": 0.12156649442480283,
      "grad_norm": 0.50390625,
      "learning_rate": 1.9971877236803336e-05,
      "loss": 0.8402,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 894
    },
    {
      "epoch": 0.12170247484362252,
      "grad_norm": 0.421875,
      "learning_rate": 1.997152029552923e-05,
      "loss": 0.6509,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 895
    },
    {
      "epoch": 0.1218384552624422,
      "grad_norm": 0.5625,
      "learning_rate": 1.9971161106552543e-05,
      "loss": 0.8314,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 896
    },
    {
      "epoch": 0.1219744356812619,
      "grad_norm": 0.416015625,
      "learning_rate": 1.9970799669954227e-05,
      "loss": 0.7559,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 897
    },
    {
      "epoch": 0.12211041610008158,
      "grad_norm": 0.45703125,
      "learning_rate": 1.997043598581577e-05,
      "loss": 0.6233,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 898
    },
    {
      "epoch": 0.12224639651890128,
      "grad_norm": 0.400390625,
      "learning_rate": 1.9970070054219136e-05,
      "loss": 0.7186,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 899
    },
    {
      "epoch": 0.12238237693772097,
      "grad_norm": 0.46484375,
      "learning_rate": 1.9969701875246817e-05,
      "loss": 0.5802,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 900
    },
    {
      "epoch": 0.12251835735654065,
      "grad_norm": 0.36328125,
      "learning_rate": 1.996933144898181e-05,
      "loss": 0.6935,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 901
    },
    {
      "epoch": 0.12265433777536035,
      "grad_norm": 0.828125,
      "learning_rate": 1.9968958775507605e-05,
      "loss": 0.862,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 902
    },
    {
      "epoch": 0.12279031819418004,
      "grad_norm": 0.3671875,
      "learning_rate": 1.996858385490821e-05,
      "loss": 0.5695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 903
    },
    {
      "epoch": 0.12292629861299972,
      "grad_norm": 0.388671875,
      "learning_rate": 1.996820668726814e-05,
      "loss": 0.6877,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 904
    },
    {
      "epoch": 0.12306227903181942,
      "grad_norm": 0.318359375,
      "learning_rate": 1.9967827272672407e-05,
      "loss": 0.6325,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 905
    },
    {
      "epoch": 0.12319825945063911,
      "grad_norm": 0.61328125,
      "learning_rate": 1.996744561120654e-05,
      "loss": 0.6074,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 906
    },
    {
      "epoch": 0.1233342398694588,
      "grad_norm": 0.38671875,
      "learning_rate": 1.9967061702956574e-05,
      "loss": 0.5391,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 907
    },
    {
      "epoch": 0.12347022028827849,
      "grad_norm": 0.458984375,
      "learning_rate": 1.9966675548009037e-05,
      "loss": 0.5195,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 908
    },
    {
      "epoch": 0.12360620070709817,
      "grad_norm": 0.400390625,
      "learning_rate": 1.9966287146450984e-05,
      "loss": 0.6043,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 909
    },
    {
      "epoch": 0.12374218112591787,
      "grad_norm": 0.65625,
      "learning_rate": 1.996589649836995e-05,
      "loss": 0.9186,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 910
    },
    {
      "epoch": 0.12387816154473756,
      "grad_norm": 0.3125,
      "learning_rate": 1.9965503603854012e-05,
      "loss": 0.5485,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 911
    },
    {
      "epoch": 0.12401414196355724,
      "grad_norm": 0.6328125,
      "learning_rate": 1.9965108462991717e-05,
      "loss": 0.576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 912
    },
    {
      "epoch": 0.12415012238237694,
      "grad_norm": 0.51171875,
      "learning_rate": 1.996471107587214e-05,
      "loss": 0.7178,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 913
    },
    {
      "epoch": 0.12428610280119663,
      "grad_norm": 0.62890625,
      "learning_rate": 1.9964311442584857e-05,
      "loss": 0.853,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 914
    },
    {
      "epoch": 0.12442208322001631,
      "grad_norm": 0.296875,
      "learning_rate": 1.996390956321995e-05,
      "loss": 0.533,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 915
    },
    {
      "epoch": 0.12455806363883601,
      "grad_norm": 0.61328125,
      "learning_rate": 1.996350543786801e-05,
      "loss": 0.9316,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 916
    },
    {
      "epoch": 0.1246940440576557,
      "grad_norm": 0.46875,
      "learning_rate": 1.996309906662013e-05,
      "loss": 0.8706,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 917
    },
    {
      "epoch": 0.12483002447647538,
      "grad_norm": 1.0390625,
      "learning_rate": 1.996269044956791e-05,
      "loss": 0.6429,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 918
    },
    {
      "epoch": 0.12496600489529508,
      "grad_norm": 1.375,
      "learning_rate": 1.9962279586803456e-05,
      "loss": 0.5833,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 919
    },
    {
      "epoch": 0.12510198531411476,
      "grad_norm": 0.7578125,
      "learning_rate": 1.9961866478419387e-05,
      "loss": 0.4321,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 920
    },
    {
      "epoch": 0.12523796573293447,
      "grad_norm": 0.439453125,
      "learning_rate": 1.996145112450882e-05,
      "loss": 0.5547,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 921
    },
    {
      "epoch": 0.12537394615175415,
      "grad_norm": 0.33984375,
      "learning_rate": 1.9961033525165375e-05,
      "loss": 0.5241,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 922
    },
    {
      "epoch": 0.12550992657057383,
      "grad_norm": 0.49609375,
      "learning_rate": 1.9960613680483192e-05,
      "loss": 0.7559,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 923
    },
    {
      "epoch": 0.12564590698939354,
      "grad_norm": 0.41015625,
      "learning_rate": 1.9960191590556905e-05,
      "loss": 0.6034,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 924
    },
    {
      "epoch": 0.12578188740821322,
      "grad_norm": 1.421875,
      "learning_rate": 1.9959767255481662e-05,
      "loss": 0.7069,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 925
    },
    {
      "epoch": 0.1259178678270329,
      "grad_norm": 0.8125,
      "learning_rate": 1.995934067535311e-05,
      "loss": 0.6419,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 926
    },
    {
      "epoch": 0.12605384824585258,
      "grad_norm": 0.365234375,
      "learning_rate": 1.9958911850267403e-05,
      "loss": 0.5957,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 927
    },
    {
      "epoch": 0.1261898286646723,
      "grad_norm": 0.296875,
      "learning_rate": 1.995848078032121e-05,
      "loss": 0.4338,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 928
    },
    {
      "epoch": 0.12632580908349197,
      "grad_norm": 0.4296875,
      "learning_rate": 1.9958047465611693e-05,
      "loss": 0.6878,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 929
    },
    {
      "epoch": 0.12646178950231166,
      "grad_norm": 0.53125,
      "learning_rate": 1.995761190623653e-05,
      "loss": 0.6214,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 930
    },
    {
      "epoch": 0.12659776992113136,
      "grad_norm": 0.51953125,
      "learning_rate": 1.99571741022939e-05,
      "loss": 0.7498,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 931
    },
    {
      "epoch": 0.12673375033995105,
      "grad_norm": 0.384765625,
      "learning_rate": 1.9956734053882494e-05,
      "loss": 0.6637,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 932
    },
    {
      "epoch": 0.12686973075877073,
      "grad_norm": 0.60546875,
      "learning_rate": 1.9956291761101498e-05,
      "loss": 0.7935,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 933
    },
    {
      "epoch": 0.12700571117759044,
      "grad_norm": 0.3125,
      "learning_rate": 1.9955847224050614e-05,
      "loss": 0.5706,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 934
    },
    {
      "epoch": 0.12714169159641012,
      "grad_norm": 0.416015625,
      "learning_rate": 1.9955400442830043e-05,
      "loss": 0.5934,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 935
    },
    {
      "epoch": 0.1272776720152298,
      "grad_norm": 0.63671875,
      "learning_rate": 1.9954951417540494e-05,
      "loss": 0.5667,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 936
    },
    {
      "epoch": 0.1274136524340495,
      "grad_norm": 0.703125,
      "learning_rate": 1.995450014828319e-05,
      "loss": 0.6117,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 937
    },
    {
      "epoch": 0.1275496328528692,
      "grad_norm": 0.45703125,
      "learning_rate": 1.9954046635159848e-05,
      "loss": 0.6524,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 938
    },
    {
      "epoch": 0.12768561327168887,
      "grad_norm": 0.48046875,
      "learning_rate": 1.9953590878272692e-05,
      "loss": 0.755,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 939
    },
    {
      "epoch": 0.12782159369050858,
      "grad_norm": 0.38671875,
      "learning_rate": 1.9953132877724462e-05,
      "loss": 0.6589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 940
    },
    {
      "epoch": 0.12795757410932826,
      "grad_norm": 0.90234375,
      "learning_rate": 1.995267263361839e-05,
      "loss": 0.8628,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 941
    },
    {
      "epoch": 0.12809355452814794,
      "grad_norm": 0.40625,
      "learning_rate": 1.9952210146058227e-05,
      "loss": 0.721,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 942
    },
    {
      "epoch": 0.12822953494696765,
      "grad_norm": 0.46875,
      "learning_rate": 1.995174541514822e-05,
      "loss": 0.7957,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 943
    },
    {
      "epoch": 0.12836551536578733,
      "grad_norm": 0.37890625,
      "learning_rate": 1.9951278440993128e-05,
      "loss": 0.6125,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 944
    },
    {
      "epoch": 0.128501495784607,
      "grad_norm": 0.369140625,
      "learning_rate": 1.995080922369821e-05,
      "loss": 0.5758,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 945
    },
    {
      "epoch": 0.1286374762034267,
      "grad_norm": 0.326171875,
      "learning_rate": 1.995033776336923e-05,
      "loss": 0.5921,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 946
    },
    {
      "epoch": 0.1287734566222464,
      "grad_norm": 0.40234375,
      "learning_rate": 1.9949864060112466e-05,
      "loss": 0.6081,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 947
    },
    {
      "epoch": 0.12890943704106608,
      "grad_norm": 0.361328125,
      "learning_rate": 1.9949388114034693e-05,
      "loss": 0.4583,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 948
    },
    {
      "epoch": 0.12904541745988576,
      "grad_norm": 0.33203125,
      "learning_rate": 1.9948909925243202e-05,
      "loss": 0.4806,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 949
    },
    {
      "epoch": 0.12918139787870547,
      "grad_norm": 1.2890625,
      "learning_rate": 1.9948429493845775e-05,
      "loss": 0.6196,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 950
    },
    {
      "epoch": 0.12931737829752515,
      "grad_norm": 0.4375,
      "learning_rate": 1.9947946819950707e-05,
      "loss": 0.7553,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 951
    },
    {
      "epoch": 0.12945335871634484,
      "grad_norm": 0.51953125,
      "learning_rate": 1.9947461903666805e-05,
      "loss": 0.818,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 952
    },
    {
      "epoch": 0.12958933913516454,
      "grad_norm": 0.431640625,
      "learning_rate": 1.994697474510337e-05,
      "loss": 0.918,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 953
    },
    {
      "epoch": 0.12972531955398423,
      "grad_norm": 0.3828125,
      "learning_rate": 1.994648534437022e-05,
      "loss": 0.6901,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 954
    },
    {
      "epoch": 0.1298612999728039,
      "grad_norm": 0.48828125,
      "learning_rate": 1.994599370157766e-05,
      "loss": 0.5968,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 955
    },
    {
      "epoch": 0.12999728039162362,
      "grad_norm": 0.50390625,
      "learning_rate": 1.994549981683652e-05,
      "loss": 0.7632,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 956
    },
    {
      "epoch": 0.1301332608104433,
      "grad_norm": 0.72265625,
      "learning_rate": 1.9945003690258127e-05,
      "loss": 0.6034,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 957
    },
    {
      "epoch": 0.13026924122926298,
      "grad_norm": 0.4296875,
      "learning_rate": 1.9944505321954315e-05,
      "loss": 0.743,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 958
    },
    {
      "epoch": 0.1304052216480827,
      "grad_norm": 0.515625,
      "learning_rate": 1.9944004712037422e-05,
      "loss": 0.6491,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 959
    },
    {
      "epoch": 0.13054120206690237,
      "grad_norm": 0.388671875,
      "learning_rate": 1.994350186062029e-05,
      "loss": 0.6207,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 960
    },
    {
      "epoch": 0.13067718248572205,
      "grad_norm": 0.408203125,
      "learning_rate": 1.994299676781627e-05,
      "loss": 0.6641,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 961
    },
    {
      "epoch": 0.13081316290454176,
      "grad_norm": 0.65234375,
      "learning_rate": 1.994248943373921e-05,
      "loss": 0.8486,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 962
    },
    {
      "epoch": 0.13094914332336144,
      "grad_norm": 0.55859375,
      "learning_rate": 1.994197985850348e-05,
      "loss": 0.8011,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 963
    },
    {
      "epoch": 0.13108512374218112,
      "grad_norm": 0.65625,
      "learning_rate": 1.9941468042223936e-05,
      "loss": 0.763,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 964
    },
    {
      "epoch": 0.13122110416100083,
      "grad_norm": 0.451171875,
      "learning_rate": 1.994095398501595e-05,
      "loss": 0.6084,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 965
    },
    {
      "epoch": 0.1313570845798205,
      "grad_norm": 0.734375,
      "learning_rate": 1.9940437686995394e-05,
      "loss": 0.8003,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 966
    },
    {
      "epoch": 0.1314930649986402,
      "grad_norm": 0.345703125,
      "learning_rate": 1.9939919148278654e-05,
      "loss": 0.4635,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 967
    },
    {
      "epoch": 0.13162904541745987,
      "grad_norm": 0.4453125,
      "learning_rate": 1.993939836898261e-05,
      "loss": 0.6427,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 968
    },
    {
      "epoch": 0.13176502583627958,
      "grad_norm": 0.392578125,
      "learning_rate": 1.9938875349224655e-05,
      "loss": 0.6338,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 969
    },
    {
      "epoch": 0.13190100625509926,
      "grad_norm": 0.4609375,
      "learning_rate": 1.993835008912268e-05,
      "loss": 0.6769,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 970
    },
    {
      "epoch": 0.13203698667391894,
      "grad_norm": 0.42578125,
      "learning_rate": 1.9937822588795096e-05,
      "loss": 0.8145,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 971
    },
    {
      "epoch": 0.13217296709273865,
      "grad_norm": 1.109375,
      "learning_rate": 1.993729284836079e-05,
      "loss": 0.8276,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 972
    },
    {
      "epoch": 0.13230894751155833,
      "grad_norm": 0.439453125,
      "learning_rate": 1.993676086793919e-05,
      "loss": 0.7707,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 973
    },
    {
      "epoch": 0.13244492793037801,
      "grad_norm": 0.4453125,
      "learning_rate": 1.99362266476502e-05,
      "loss": 0.601,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 974
    },
    {
      "epoch": 0.13258090834919772,
      "grad_norm": 0.322265625,
      "learning_rate": 1.993569018761424e-05,
      "loss": 0.5931,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 975
    },
    {
      "epoch": 0.1327168887680174,
      "grad_norm": 0.6015625,
      "learning_rate": 1.993515148795224e-05,
      "loss": 0.513,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 976
    },
    {
      "epoch": 0.13285286918683709,
      "grad_norm": 0.43359375,
      "learning_rate": 1.993461054878562e-05,
      "loss": 0.5488,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 977
    },
    {
      "epoch": 0.1329888496056568,
      "grad_norm": 0.5234375,
      "learning_rate": 1.993406737023633e-05,
      "loss": 0.3774,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 978
    },
    {
      "epoch": 0.13312483002447648,
      "grad_norm": 0.33203125,
      "learning_rate": 1.9933521952426796e-05,
      "loss": 0.6198,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 979
    },
    {
      "epoch": 0.13326081044329616,
      "grad_norm": 0.44140625,
      "learning_rate": 1.9932974295479965e-05,
      "loss": 0.8618,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 980
    },
    {
      "epoch": 0.13339679086211587,
      "grad_norm": 0.54296875,
      "learning_rate": 1.9932424399519286e-05,
      "loss": 0.9338,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 981
    },
    {
      "epoch": 0.13353277128093555,
      "grad_norm": 0.31640625,
      "learning_rate": 1.9931872264668717e-05,
      "loss": 0.3941,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 982
    },
    {
      "epoch": 0.13366875169975523,
      "grad_norm": 0.38671875,
      "learning_rate": 1.9931317891052707e-05,
      "loss": 0.639,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 983
    },
    {
      "epoch": 0.13380473211857494,
      "grad_norm": 0.43359375,
      "learning_rate": 1.9930761278796228e-05,
      "loss": 0.6043,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 984
    },
    {
      "epoch": 0.13394071253739462,
      "grad_norm": 0.4921875,
      "learning_rate": 1.993020242802474e-05,
      "loss": 0.508,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 985
    },
    {
      "epoch": 0.1340766929562143,
      "grad_norm": 0.396484375,
      "learning_rate": 1.9929641338864213e-05,
      "loss": 0.7293,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 986
    },
    {
      "epoch": 0.134212673375034,
      "grad_norm": 0.3984375,
      "learning_rate": 1.9929078011441133e-05,
      "loss": 0.6965,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 987
    },
    {
      "epoch": 0.1343486537938537,
      "grad_norm": 0.427734375,
      "learning_rate": 1.9928512445882474e-05,
      "loss": 0.6818,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 988
    },
    {
      "epoch": 0.13448463421267337,
      "grad_norm": 0.52734375,
      "learning_rate": 1.9927944642315726e-05,
      "loss": 0.6385,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 989
    },
    {
      "epoch": 0.13462061463149305,
      "grad_norm": 0.6328125,
      "learning_rate": 1.9927374600868873e-05,
      "loss": 0.7975,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 990
    },
    {
      "epoch": 0.13475659505031276,
      "grad_norm": 0.5,
      "learning_rate": 1.992680232167041e-05,
      "loss": 0.6315,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 991
    },
    {
      "epoch": 0.13489257546913244,
      "grad_norm": 0.380859375,
      "learning_rate": 1.992622780484934e-05,
      "loss": 0.6209,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 992
    },
    {
      "epoch": 0.13502855588795212,
      "grad_norm": 0.52734375,
      "learning_rate": 1.9925651050535165e-05,
      "loss": 0.6364,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 993
    },
    {
      "epoch": 0.13516453630677183,
      "grad_norm": 0.384765625,
      "learning_rate": 1.9925072058857887e-05,
      "loss": 0.8385,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 994
    },
    {
      "epoch": 0.1353005167255915,
      "grad_norm": 0.97265625,
      "learning_rate": 1.992449082994803e-05,
      "loss": 0.688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 995
    },
    {
      "epoch": 0.1354364971444112,
      "grad_norm": 0.271484375,
      "learning_rate": 1.9923907363936593e-05,
      "loss": 0.4967,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 996
    },
    {
      "epoch": 0.1355724775632309,
      "grad_norm": 1.046875,
      "learning_rate": 1.9923321660955113e-05,
      "loss": 0.9671,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 997
    },
    {
      "epoch": 0.13570845798205058,
      "grad_norm": 0.330078125,
      "learning_rate": 1.9922733721135606e-05,
      "loss": 0.5848,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 998
    },
    {
      "epoch": 0.13584443840087027,
      "grad_norm": 0.328125,
      "learning_rate": 1.99221435446106e-05,
      "loss": 0.5133,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 999
    },
    {
      "epoch": 0.13598041881968997,
      "grad_norm": 0.451171875,
      "learning_rate": 1.992155113151313e-05,
      "loss": 0.7087,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1000
    },
    {
      "epoch": 0.13611639923850966,
      "grad_norm": 4.15625,
      "learning_rate": 1.9920956481976734e-05,
      "loss": 0.6852,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1001
    },
    {
      "epoch": 0.13625237965732934,
      "grad_norm": 0.65234375,
      "learning_rate": 1.9920359596135454e-05,
      "loss": 0.5228,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1002
    },
    {
      "epoch": 0.13638836007614905,
      "grad_norm": 0.41015625,
      "learning_rate": 1.9919760474123832e-05,
      "loss": 0.8451,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1003
    },
    {
      "epoch": 0.13652434049496873,
      "grad_norm": 0.41796875,
      "learning_rate": 1.991915911607692e-05,
      "loss": 0.5782,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1004
    },
    {
      "epoch": 0.1366603209137884,
      "grad_norm": 0.53515625,
      "learning_rate": 1.991855552213027e-05,
      "loss": 0.7952,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1005
    },
    {
      "epoch": 0.13679630133260812,
      "grad_norm": 0.44921875,
      "learning_rate": 1.9917949692419943e-05,
      "loss": 0.5933,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1006
    },
    {
      "epoch": 0.1369322817514278,
      "grad_norm": 0.384765625,
      "learning_rate": 1.9917341627082494e-05,
      "loss": 0.4313,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1007
    },
    {
      "epoch": 0.13706826217024748,
      "grad_norm": 0.546875,
      "learning_rate": 1.9916731326254997e-05,
      "loss": 0.7194,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1008
    },
    {
      "epoch": 0.1372042425890672,
      "grad_norm": 0.4296875,
      "learning_rate": 1.991611879007501e-05,
      "loss": 0.7493,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1009
    },
    {
      "epoch": 0.13734022300788687,
      "grad_norm": 0.322265625,
      "learning_rate": 1.9915504018680617e-05,
      "loss": 0.554,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1010
    },
    {
      "epoch": 0.13747620342670655,
      "grad_norm": 0.58203125,
      "learning_rate": 1.991488701221039e-05,
      "loss": 0.6828,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1011
    },
    {
      "epoch": 0.13761218384552623,
      "grad_norm": 0.35546875,
      "learning_rate": 1.9914267770803407e-05,
      "loss": 0.5615,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1012
    },
    {
      "epoch": 0.13774816426434594,
      "grad_norm": 0.55859375,
      "learning_rate": 1.991364629459926e-05,
      "loss": 0.6545,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1013
    },
    {
      "epoch": 0.13788414468316562,
      "grad_norm": 0.279296875,
      "learning_rate": 1.9913022583738032e-05,
      "loss": 0.3615,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1014
    },
    {
      "epoch": 0.1380201251019853,
      "grad_norm": 0.70703125,
      "learning_rate": 1.9912396638360314e-05,
      "loss": 0.6746,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1015
    },
    {
      "epoch": 0.138156105520805,
      "grad_norm": 0.51171875,
      "learning_rate": 1.9911768458607206e-05,
      "loss": 0.5874,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1016
    },
    {
      "epoch": 0.1382920859396247,
      "grad_norm": 0.337890625,
      "learning_rate": 1.9911138044620304e-05,
      "loss": 0.541,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1017
    },
    {
      "epoch": 0.13842806635844437,
      "grad_norm": 0.298828125,
      "learning_rate": 1.9910505396541712e-05,
      "loss": 0.5576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1018
    },
    {
      "epoch": 0.13856404677726408,
      "grad_norm": 0.2890625,
      "learning_rate": 1.990987051451404e-05,
      "loss": 0.4857,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1019
    },
    {
      "epoch": 0.13870002719608376,
      "grad_norm": 0.302734375,
      "learning_rate": 1.9909233398680393e-05,
      "loss": 0.6004,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1020
    },
    {
      "epoch": 0.13883600761490344,
      "grad_norm": 0.375,
      "learning_rate": 1.990859404918439e-05,
      "loss": 0.6766,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1021
    },
    {
      "epoch": 0.13897198803372315,
      "grad_norm": 0.3125,
      "learning_rate": 1.990795246617014e-05,
      "loss": 0.5635,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1022
    },
    {
      "epoch": 0.13910796845254284,
      "grad_norm": 0.296875,
      "learning_rate": 1.990730864978227e-05,
      "loss": 0.5182,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1023
    },
    {
      "epoch": 0.13924394887136252,
      "grad_norm": 0.38671875,
      "learning_rate": 1.9906662600165906e-05,
      "loss": 0.7386,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1024
    },
    {
      "epoch": 0.13937992929018223,
      "grad_norm": 0.59765625,
      "learning_rate": 1.990601431746667e-05,
      "loss": 0.6359,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1025
    },
    {
      "epoch": 0.1395159097090019,
      "grad_norm": 0.462890625,
      "learning_rate": 1.9905363801830695e-05,
      "loss": 0.8268,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1026
    },
    {
      "epoch": 0.1396518901278216,
      "grad_norm": 0.3984375,
      "learning_rate": 1.9904711053404614e-05,
      "loss": 0.6481,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1027
    },
    {
      "epoch": 0.1397878705466413,
      "grad_norm": 0.3359375,
      "learning_rate": 1.9904056072335568e-05,
      "loss": 0.6668,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1028
    },
    {
      "epoch": 0.13992385096546098,
      "grad_norm": 0.51171875,
      "learning_rate": 1.9903398858771195e-05,
      "loss": 0.7178,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1029
    },
    {
      "epoch": 0.14005983138428066,
      "grad_norm": 0.5546875,
      "learning_rate": 1.990273941285964e-05,
      "loss": 0.7139,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1030
    },
    {
      "epoch": 0.14019581180310034,
      "grad_norm": 0.6484375,
      "learning_rate": 1.990207773474955e-05,
      "loss": 0.6335,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1031
    },
    {
      "epoch": 0.14033179222192005,
      "grad_norm": 0.44140625,
      "learning_rate": 1.9901413824590072e-05,
      "loss": 0.6893,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1032
    },
    {
      "epoch": 0.14046777264073973,
      "grad_norm": 0.5625,
      "learning_rate": 1.9900747682530864e-05,
      "loss": 0.731,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1033
    },
    {
      "epoch": 0.1406037530595594,
      "grad_norm": 0.478515625,
      "learning_rate": 1.990007930872208e-05,
      "loss": 0.7915,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1034
    },
    {
      "epoch": 0.14073973347837912,
      "grad_norm": 0.54296875,
      "learning_rate": 1.9899408703314383e-05,
      "loss": 0.7218,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1035
    },
    {
      "epoch": 0.1408757138971988,
      "grad_norm": 0.4375,
      "learning_rate": 1.9898735866458935e-05,
      "loss": 0.5785,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1036
    },
    {
      "epoch": 0.14101169431601848,
      "grad_norm": 0.3359375,
      "learning_rate": 1.9898060798307396e-05,
      "loss": 0.6484,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1037
    },
    {
      "epoch": 0.1411476747348382,
      "grad_norm": 0.34765625,
      "learning_rate": 1.989738349901194e-05,
      "loss": 0.6992,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1038
    },
    {
      "epoch": 0.14128365515365787,
      "grad_norm": 0.423828125,
      "learning_rate": 1.9896703968725236e-05,
      "loss": 0.6302,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1039
    },
    {
      "epoch": 0.14141963557247755,
      "grad_norm": 0.515625,
      "learning_rate": 1.989602220760046e-05,
      "loss": 0.5436,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1040
    },
    {
      "epoch": 0.14155561599129726,
      "grad_norm": 0.328125,
      "learning_rate": 1.9895338215791292e-05,
      "loss": 0.5828,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1041
    },
    {
      "epoch": 0.14169159641011694,
      "grad_norm": 0.67578125,
      "learning_rate": 1.9894651993451907e-05,
      "loss": 0.6473,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1042
    },
    {
      "epoch": 0.14182757682893662,
      "grad_norm": 0.42578125,
      "learning_rate": 1.989396354073699e-05,
      "loss": 0.7012,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1043
    },
    {
      "epoch": 0.14196355724775633,
      "grad_norm": 0.2734375,
      "learning_rate": 1.9893272857801726e-05,
      "loss": 0.4551,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1044
    },
    {
      "epoch": 0.14209953766657601,
      "grad_norm": 0.416015625,
      "learning_rate": 1.9892579944801806e-05,
      "loss": 0.668,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1045
    },
    {
      "epoch": 0.1422355180853957,
      "grad_norm": 0.3125,
      "learning_rate": 1.9891884801893414e-05,
      "loss": 0.6064,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1046
    },
    {
      "epoch": 0.1423714985042154,
      "grad_norm": 0.267578125,
      "learning_rate": 1.9891187429233253e-05,
      "loss": 0.5319,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1047
    },
    {
      "epoch": 0.14250747892303509,
      "grad_norm": 0.2890625,
      "learning_rate": 1.989048782697851e-05,
      "loss": 0.3743,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1048
    },
    {
      "epoch": 0.14264345934185477,
      "grad_norm": 0.6796875,
      "learning_rate": 1.9889785995286894e-05,
      "loss": 0.6494,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1049
    },
    {
      "epoch": 0.14277943976067448,
      "grad_norm": 1.03125,
      "learning_rate": 1.98890819343166e-05,
      "loss": 0.8675,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1050
    },
    {
      "epoch": 0.14291542017949416,
      "grad_norm": 0.400390625,
      "learning_rate": 1.9888375644226333e-05,
      "loss": 0.5254,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1051
    },
    {
      "epoch": 0.14305140059831384,
      "grad_norm": 0.345703125,
      "learning_rate": 1.98876671251753e-05,
      "loss": 0.541,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1052
    },
    {
      "epoch": 0.14318738101713352,
      "grad_norm": 0.7734375,
      "learning_rate": 1.9886956377323207e-05,
      "loss": 0.7489,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1053
    },
    {
      "epoch": 0.14332336143595323,
      "grad_norm": 0.388671875,
      "learning_rate": 1.9886243400830272e-05,
      "loss": 0.4702,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1054
    },
    {
      "epoch": 0.1434593418547729,
      "grad_norm": 0.55078125,
      "learning_rate": 1.9885528195857204e-05,
      "loss": 0.7684,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1055
    },
    {
      "epoch": 0.1435953222735926,
      "grad_norm": 0.50390625,
      "learning_rate": 1.9884810762565216e-05,
      "loss": 0.6664,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1056
    },
    {
      "epoch": 0.1437313026924123,
      "grad_norm": 1.1015625,
      "learning_rate": 1.9884091101116032e-05,
      "loss": 0.9056,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1057
    },
    {
      "epoch": 0.14386728311123198,
      "grad_norm": 0.58984375,
      "learning_rate": 1.988336921167187e-05,
      "loss": 0.6343,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1058
    },
    {
      "epoch": 0.14400326353005166,
      "grad_norm": 0.51953125,
      "learning_rate": 1.988264509439545e-05,
      "loss": 0.6582,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1059
    },
    {
      "epoch": 0.14413924394887137,
      "grad_norm": 0.375,
      "learning_rate": 1.9881918749450004e-05,
      "loss": 0.5641,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1060
    },
    {
      "epoch": 0.14427522436769105,
      "grad_norm": 0.51171875,
      "learning_rate": 1.9881190176999255e-05,
      "loss": 0.7757,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1061
    },
    {
      "epoch": 0.14441120478651073,
      "grad_norm": 0.34765625,
      "learning_rate": 1.988045937720743e-05,
      "loss": 0.5729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1062
    },
    {
      "epoch": 0.14454718520533044,
      "grad_norm": 0.400390625,
      "learning_rate": 1.987972635023926e-05,
      "loss": 0.7183,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1063
    },
    {
      "epoch": 0.14468316562415012,
      "grad_norm": 0.33203125,
      "learning_rate": 1.9878991096259985e-05,
      "loss": 0.466,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1064
    },
    {
      "epoch": 0.1448191460429698,
      "grad_norm": 0.65234375,
      "learning_rate": 1.987825361543533e-05,
      "loss": 0.6888,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1065
    },
    {
      "epoch": 0.1449551264617895,
      "grad_norm": 0.40625,
      "learning_rate": 1.9877513907931543e-05,
      "loss": 0.7483,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1066
    },
    {
      "epoch": 0.1450911068806092,
      "grad_norm": 0.50390625,
      "learning_rate": 1.9876771973915355e-05,
      "loss": 0.8032,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1067
    },
    {
      "epoch": 0.14522708729942888,
      "grad_norm": 0.3203125,
      "learning_rate": 1.987602781355401e-05,
      "loss": 0.5272,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1068
    },
    {
      "epoch": 0.14536306771824858,
      "grad_norm": 0.263671875,
      "learning_rate": 1.9875281427015252e-05,
      "loss": 0.4329,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1069
    },
    {
      "epoch": 0.14549904813706827,
      "grad_norm": 0.5546875,
      "learning_rate": 1.987453281446732e-05,
      "loss": 0.5986,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1070
    },
    {
      "epoch": 0.14563502855588795,
      "grad_norm": 0.380859375,
      "learning_rate": 1.987378197607897e-05,
      "loss": 0.6388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1071
    },
    {
      "epoch": 0.14577100897470766,
      "grad_norm": 0.32421875,
      "learning_rate": 1.9873028912019444e-05,
      "loss": 0.6108,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1072
    },
    {
      "epoch": 0.14590698939352734,
      "grad_norm": 0.251953125,
      "learning_rate": 1.9872273622458494e-05,
      "loss": 0.4535,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1073
    },
    {
      "epoch": 0.14604296981234702,
      "grad_norm": 0.3828125,
      "learning_rate": 1.9871516107566366e-05,
      "loss": 0.5742,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1074
    },
    {
      "epoch": 0.1461789502311667,
      "grad_norm": 0.32421875,
      "learning_rate": 1.987075636751382e-05,
      "loss": 0.5293,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1075
    },
    {
      "epoch": 0.1463149306499864,
      "grad_norm": 0.27734375,
      "learning_rate": 1.9869994402472108e-05,
      "loss": 0.5568,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1076
    },
    {
      "epoch": 0.1464509110688061,
      "grad_norm": 0.408203125,
      "learning_rate": 1.9869230212612986e-05,
      "loss": 0.5788,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1077
    },
    {
      "epoch": 0.14658689148762577,
      "grad_norm": 0.44140625,
      "learning_rate": 1.9868463798108714e-05,
      "loss": 0.54,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1078
    },
    {
      "epoch": 0.14672287190644548,
      "grad_norm": 0.7890625,
      "learning_rate": 1.9867695159132045e-05,
      "loss": 0.6463,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1079
    },
    {
      "epoch": 0.14685885232526516,
      "grad_norm": 0.41015625,
      "learning_rate": 1.9866924295856245e-05,
      "loss": 0.6922,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1080
    },
    {
      "epoch": 0.14699483274408484,
      "grad_norm": 0.296875,
      "learning_rate": 1.986615120845508e-05,
      "loss": 0.5762,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1081
    },
    {
      "epoch": 0.14713081316290455,
      "grad_norm": 0.361328125,
      "learning_rate": 1.9865375897102806e-05,
      "loss": 0.6546,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1082
    },
    {
      "epoch": 0.14726679358172423,
      "grad_norm": 0.380859375,
      "learning_rate": 1.9864598361974184e-05,
      "loss": 0.6694,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1083
    },
    {
      "epoch": 0.1474027740005439,
      "grad_norm": 0.64453125,
      "learning_rate": 1.9863818603244492e-05,
      "loss": 0.6701,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1084
    },
    {
      "epoch": 0.14753875441936362,
      "grad_norm": 0.6640625,
      "learning_rate": 1.986303662108949e-05,
      "loss": 0.6712,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1085
    },
    {
      "epoch": 0.1476747348381833,
      "grad_norm": 0.6171875,
      "learning_rate": 1.9862252415685447e-05,
      "loss": 0.8314,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1086
    },
    {
      "epoch": 0.14781071525700298,
      "grad_norm": 0.32421875,
      "learning_rate": 1.986146598720913e-05,
      "loss": 0.507,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1087
    },
    {
      "epoch": 0.1479466956758227,
      "grad_norm": 0.392578125,
      "learning_rate": 1.9860677335837818e-05,
      "loss": 0.5885,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1088
    },
    {
      "epoch": 0.14808267609464237,
      "grad_norm": 0.44140625,
      "learning_rate": 1.9859886461749275e-05,
      "loss": 0.7633,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1089
    },
    {
      "epoch": 0.14821865651346205,
      "grad_norm": 0.41015625,
      "learning_rate": 1.9859093365121773e-05,
      "loss": 0.6082,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1090
    },
    {
      "epoch": 0.14835463693228176,
      "grad_norm": 0.625,
      "learning_rate": 1.9858298046134094e-05,
      "loss": 0.631,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1091
    },
    {
      "epoch": 0.14849061735110144,
      "grad_norm": 0.515625,
      "learning_rate": 1.9857500504965503e-05,
      "loss": 0.8205,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1092
    },
    {
      "epoch": 0.14862659776992113,
      "grad_norm": 0.37890625,
      "learning_rate": 1.9856700741795777e-05,
      "loss": 0.6898,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1093
    },
    {
      "epoch": 0.14876257818874083,
      "grad_norm": 0.4453125,
      "learning_rate": 1.9855898756805197e-05,
      "loss": 0.5993,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1094
    },
    {
      "epoch": 0.14889855860756052,
      "grad_norm": 0.734375,
      "learning_rate": 1.985509455017454e-05,
      "loss": 0.8652,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1095
    },
    {
      "epoch": 0.1490345390263802,
      "grad_norm": 0.27734375,
      "learning_rate": 1.9854288122085082e-05,
      "loss": 0.4761,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1096
    },
    {
      "epoch": 0.14917051944519988,
      "grad_norm": 0.462890625,
      "learning_rate": 1.9853479472718603e-05,
      "loss": 0.8153,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1097
    },
    {
      "epoch": 0.1493064998640196,
      "grad_norm": 0.400390625,
      "learning_rate": 1.985266860225738e-05,
      "loss": 0.59,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1098
    },
    {
      "epoch": 0.14944248028283927,
      "grad_norm": 0.59375,
      "learning_rate": 1.9851855510884194e-05,
      "loss": 0.4793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1099
    },
    {
      "epoch": 0.14957846070165895,
      "grad_norm": 0.375,
      "learning_rate": 1.985104019878233e-05,
      "loss": 0.5822,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1100
    },
    {
      "epoch": 0.14971444112047866,
      "grad_norm": 0.6875,
      "learning_rate": 1.9850222666135558e-05,
      "loss": 0.8162,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1101
    },
    {
      "epoch": 0.14985042153929834,
      "grad_norm": 0.33203125,
      "learning_rate": 1.984940291312817e-05,
      "loss": 0.5589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1102
    },
    {
      "epoch": 0.14998640195811802,
      "grad_norm": 0.81640625,
      "learning_rate": 1.984858093994495e-05,
      "loss": 0.7357,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1103
    },
    {
      "epoch": 0.15012238237693773,
      "grad_norm": 0.6796875,
      "learning_rate": 1.9847756746771175e-05,
      "loss": 0.8205,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1104
    },
    {
      "epoch": 0.1502583627957574,
      "grad_norm": 0.431640625,
      "learning_rate": 1.984693033379263e-05,
      "loss": 0.709,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1105
    },
    {
      "epoch": 0.1503943432145771,
      "grad_norm": 0.328125,
      "learning_rate": 1.9846101701195594e-05,
      "loss": 0.6652,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1106
    },
    {
      "epoch": 0.1505303236333968,
      "grad_norm": 0.3671875,
      "learning_rate": 1.9845270849166856e-05,
      "loss": 0.5422,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1107
    },
    {
      "epoch": 0.15066630405221648,
      "grad_norm": 0.408203125,
      "learning_rate": 1.9844437777893707e-05,
      "loss": 0.6642,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1108
    },
    {
      "epoch": 0.15080228447103616,
      "grad_norm": 0.6640625,
      "learning_rate": 1.9843602487563918e-05,
      "loss": 0.5658,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1109
    },
    {
      "epoch": 0.15093826488985587,
      "grad_norm": 0.3828125,
      "learning_rate": 1.984276497836578e-05,
      "loss": 0.6483,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1110
    },
    {
      "epoch": 0.15107424530867555,
      "grad_norm": 0.353515625,
      "learning_rate": 1.984192525048808e-05,
      "loss": 0.6112,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1111
    },
    {
      "epoch": 0.15121022572749523,
      "grad_norm": 0.265625,
      "learning_rate": 1.9841083304120097e-05,
      "loss": 0.4655,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1112
    },
    {
      "epoch": 0.15134620614631494,
      "grad_norm": 0.361328125,
      "learning_rate": 1.9840239139451622e-05,
      "loss": 0.6156,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1113
    },
    {
      "epoch": 0.15148218656513462,
      "grad_norm": 1.0078125,
      "learning_rate": 1.9839392756672936e-05,
      "loss": 0.8356,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1114
    },
    {
      "epoch": 0.1516181669839543,
      "grad_norm": 0.31640625,
      "learning_rate": 1.9838544155974822e-05,
      "loss": 0.4772,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1115
    },
    {
      "epoch": 0.151754147402774,
      "grad_norm": 0.423828125,
      "learning_rate": 1.9837693337548575e-05,
      "loss": 0.6364,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1116
    },
    {
      "epoch": 0.1518901278215937,
      "grad_norm": 0.82421875,
      "learning_rate": 1.983684030158597e-05,
      "loss": 0.7549,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1117
    },
    {
      "epoch": 0.15202610824041338,
      "grad_norm": 0.60546875,
      "learning_rate": 1.9835985048279296e-05,
      "loss": 0.6429,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1118
    },
    {
      "epoch": 0.15216208865923306,
      "grad_norm": 0.392578125,
      "learning_rate": 1.983512757782134e-05,
      "loss": 0.6309,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1119
    },
    {
      "epoch": 0.15229806907805277,
      "grad_norm": 0.431640625,
      "learning_rate": 1.9834267890405375e-05,
      "loss": 0.7669,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1120
    },
    {
      "epoch": 0.15243404949687245,
      "grad_norm": 0.3515625,
      "learning_rate": 1.98334059862252e-05,
      "loss": 0.5456,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1121
    },
    {
      "epoch": 0.15257002991569213,
      "grad_norm": 0.609375,
      "learning_rate": 1.9832541865475086e-05,
      "loss": 0.8426,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1122
    },
    {
      "epoch": 0.15270601033451184,
      "grad_norm": 0.318359375,
      "learning_rate": 1.9831675528349825e-05,
      "loss": 0.5942,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1123
    },
    {
      "epoch": 0.15284199075333152,
      "grad_norm": 0.71875,
      "learning_rate": 1.9830806975044697e-05,
      "loss": 0.8075,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1124
    },
    {
      "epoch": 0.1529779711721512,
      "grad_norm": 0.36328125,
      "learning_rate": 1.9829936205755484e-05,
      "loss": 0.548,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1125
    },
    {
      "epoch": 0.1531139515909709,
      "grad_norm": 0.30859375,
      "learning_rate": 1.982906322067847e-05,
      "loss": 0.542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1126
    },
    {
      "epoch": 0.1532499320097906,
      "grad_norm": 0.416015625,
      "learning_rate": 1.9828188020010433e-05,
      "loss": 0.7406,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1127
    },
    {
      "epoch": 0.15338591242861027,
      "grad_norm": 0.69140625,
      "learning_rate": 1.982731060394866e-05,
      "loss": 0.661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1128
    },
    {
      "epoch": 0.15352189284742998,
      "grad_norm": 0.359375,
      "learning_rate": 1.9826430972690924e-05,
      "loss": 0.5698,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1129
    },
    {
      "epoch": 0.15365787326624966,
      "grad_norm": 0.7109375,
      "learning_rate": 1.982554912643551e-05,
      "loss": 0.767,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1130
    },
    {
      "epoch": 0.15379385368506934,
      "grad_norm": 0.396484375,
      "learning_rate": 1.9824665065381194e-05,
      "loss": 0.6672,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1131
    },
    {
      "epoch": 0.15392983410388905,
      "grad_norm": 0.38671875,
      "learning_rate": 1.9823778789727253e-05,
      "loss": 0.5986,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1132
    },
    {
      "epoch": 0.15406581452270873,
      "grad_norm": 0.609375,
      "learning_rate": 1.9822890299673466e-05,
      "loss": 0.6289,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1133
    },
    {
      "epoch": 0.15420179494152841,
      "grad_norm": 0.400390625,
      "learning_rate": 1.9821999595420114e-05,
      "loss": 0.7399,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1134
    },
    {
      "epoch": 0.15433777536034812,
      "grad_norm": 0.44921875,
      "learning_rate": 1.9821106677167966e-05,
      "loss": 0.668,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1135
    },
    {
      "epoch": 0.1544737557791678,
      "grad_norm": 0.466796875,
      "learning_rate": 1.9820211545118303e-05,
      "loss": 0.8298,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1136
    },
    {
      "epoch": 0.15460973619798749,
      "grad_norm": 0.76171875,
      "learning_rate": 1.981931419947289e-05,
      "loss": 0.9217,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1137
    },
    {
      "epoch": 0.15474571661680717,
      "grad_norm": 0.408203125,
      "learning_rate": 1.981841464043401e-05,
      "loss": 0.5116,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1138
    },
    {
      "epoch": 0.15488169703562688,
      "grad_norm": 0.3984375,
      "learning_rate": 1.9817512868204425e-05,
      "loss": 0.7804,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1139
    },
    {
      "epoch": 0.15501767745444656,
      "grad_norm": 0.48046875,
      "learning_rate": 1.981660888298741e-05,
      "loss": 0.6471,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1140
    },
    {
      "epoch": 0.15515365787326624,
      "grad_norm": 0.486328125,
      "learning_rate": 1.9815702684986738e-05,
      "loss": 0.5968,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1141
    },
    {
      "epoch": 0.15528963829208595,
      "grad_norm": 0.48046875,
      "learning_rate": 1.981479427440667e-05,
      "loss": 0.5347,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1142
    },
    {
      "epoch": 0.15542561871090563,
      "grad_norm": 0.263671875,
      "learning_rate": 1.9813883651451976e-05,
      "loss": 0.4196,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1143
    },
    {
      "epoch": 0.1555615991297253,
      "grad_norm": 0.396484375,
      "learning_rate": 1.981297081632792e-05,
      "loss": 0.7403,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1144
    },
    {
      "epoch": 0.15569757954854502,
      "grad_norm": 0.48046875,
      "learning_rate": 1.981205576924027e-05,
      "loss": 0.7716,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1145
    },
    {
      "epoch": 0.1558335599673647,
      "grad_norm": 0.388671875,
      "learning_rate": 1.981113851039529e-05,
      "loss": 0.7126,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1146
    },
    {
      "epoch": 0.15596954038618438,
      "grad_norm": 0.3828125,
      "learning_rate": 1.9810219039999732e-05,
      "loss": 0.6157,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1147
    },
    {
      "epoch": 0.1561055208050041,
      "grad_norm": 0.337890625,
      "learning_rate": 1.9809297358260868e-05,
      "loss": 0.5856,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1148
    },
    {
      "epoch": 0.15624150122382377,
      "grad_norm": 0.58984375,
      "learning_rate": 1.9808373465386445e-05,
      "loss": 0.7186,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1149
    },
    {
      "epoch": 0.15637748164264345,
      "grad_norm": 0.6953125,
      "learning_rate": 1.9807447361584726e-05,
      "loss": 0.7599,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1150
    },
    {
      "epoch": 0.15651346206146316,
      "grad_norm": 0.359375,
      "learning_rate": 1.9806519047064467e-05,
      "loss": 0.6315,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1151
    },
    {
      "epoch": 0.15664944248028284,
      "grad_norm": 0.9296875,
      "learning_rate": 1.980558852203492e-05,
      "loss": 0.8918,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1152
    },
    {
      "epoch": 0.15678542289910252,
      "grad_norm": 0.81640625,
      "learning_rate": 1.980465578670583e-05,
      "loss": 0.8765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1153
    },
    {
      "epoch": 0.15692140331792223,
      "grad_norm": 0.287109375,
      "learning_rate": 1.9803720841287458e-05,
      "loss": 0.4538,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1154
    },
    {
      "epoch": 0.1570573837367419,
      "grad_norm": 0.462890625,
      "learning_rate": 1.980278368599055e-05,
      "loss": 0.737,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1155
    },
    {
      "epoch": 0.1571933641555616,
      "grad_norm": 0.314453125,
      "learning_rate": 1.9801844321026345e-05,
      "loss": 0.5021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1156
    },
    {
      "epoch": 0.1573293445743813,
      "grad_norm": 0.421875,
      "learning_rate": 1.9800902746606594e-05,
      "loss": 0.6737,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1157
    },
    {
      "epoch": 0.15746532499320098,
      "grad_norm": 0.49609375,
      "learning_rate": 1.9799958962943536e-05,
      "loss": 0.751,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1158
    },
    {
      "epoch": 0.15760130541202066,
      "grad_norm": 0.703125,
      "learning_rate": 1.9799012970249913e-05,
      "loss": 0.8314,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1159
    },
    {
      "epoch": 0.15773728583084035,
      "grad_norm": 0.474609375,
      "learning_rate": 1.9798064768738965e-05,
      "loss": 0.7041,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1160
    },
    {
      "epoch": 0.15787326624966005,
      "grad_norm": 0.578125,
      "learning_rate": 1.9797114358624425e-05,
      "loss": 0.5095,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1161
    },
    {
      "epoch": 0.15800924666847974,
      "grad_norm": 0.58984375,
      "learning_rate": 1.979616174012053e-05,
      "loss": 0.6177,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1162
    },
    {
      "epoch": 0.15814522708729942,
      "grad_norm": 0.41015625,
      "learning_rate": 1.9795206913442007e-05,
      "loss": 0.6412,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1163
    },
    {
      "epoch": 0.15828120750611913,
      "grad_norm": 0.48828125,
      "learning_rate": 1.979424987880409e-05,
      "loss": 0.653,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1164
    },
    {
      "epoch": 0.1584171879249388,
      "grad_norm": 0.4375,
      "learning_rate": 1.9793290636422503e-05,
      "loss": 0.7507,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1165
    },
    {
      "epoch": 0.1585531683437585,
      "grad_norm": 0.6015625,
      "learning_rate": 1.9792329186513476e-05,
      "loss": 0.8773,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1166
    },
    {
      "epoch": 0.1586891487625782,
      "grad_norm": 0.8046875,
      "learning_rate": 1.9791365529293727e-05,
      "loss": 0.6901,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1167
    },
    {
      "epoch": 0.15882512918139788,
      "grad_norm": 0.423828125,
      "learning_rate": 1.9790399664980478e-05,
      "loss": 0.6663,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1168
    },
    {
      "epoch": 0.15896110960021756,
      "grad_norm": 0.458984375,
      "learning_rate": 1.9789431593791447e-05,
      "loss": 0.57,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1169
    },
    {
      "epoch": 0.15909709001903727,
      "grad_norm": 0.84375,
      "learning_rate": 1.9788461315944847e-05,
      "loss": 0.7396,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1170
    },
    {
      "epoch": 0.15923307043785695,
      "grad_norm": 0.38671875,
      "learning_rate": 1.9787488831659395e-05,
      "loss": 0.6574,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1171
    },
    {
      "epoch": 0.15936905085667663,
      "grad_norm": 0.2353515625,
      "learning_rate": 1.9786514141154293e-05,
      "loss": 0.4658,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1172
    },
    {
      "epoch": 0.15950503127549634,
      "grad_norm": 0.396484375,
      "learning_rate": 1.9785537244649258e-05,
      "loss": 0.5817,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1173
    },
    {
      "epoch": 0.15964101169431602,
      "grad_norm": 0.4453125,
      "learning_rate": 1.9784558142364485e-05,
      "loss": 0.8245,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1174
    },
    {
      "epoch": 0.1597769921131357,
      "grad_norm": 0.314453125,
      "learning_rate": 1.978357683452068e-05,
      "loss": 0.625,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1175
    },
    {
      "epoch": 0.1599129725319554,
      "grad_norm": 0.62890625,
      "learning_rate": 1.9782593321339045e-05,
      "loss": 0.6803,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1176
    },
    {
      "epoch": 0.1600489529507751,
      "grad_norm": 0.3046875,
      "learning_rate": 1.978160760304127e-05,
      "loss": 0.5366,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1177
    },
    {
      "epoch": 0.16018493336959477,
      "grad_norm": 0.31640625,
      "learning_rate": 1.9780619679849552e-05,
      "loss": 0.619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1178
    },
    {
      "epoch": 0.16032091378841448,
      "grad_norm": 0.6015625,
      "learning_rate": 1.977962955198658e-05,
      "loss": 0.7303,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1179
    },
    {
      "epoch": 0.16045689420723416,
      "grad_norm": 0.322265625,
      "learning_rate": 1.9778637219675537e-05,
      "loss": 0.5314,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1180
    },
    {
      "epoch": 0.16059287462605384,
      "grad_norm": 0.396484375,
      "learning_rate": 1.9777642683140112e-05,
      "loss": 0.5911,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1181
    },
    {
      "epoch": 0.16072885504487353,
      "grad_norm": 0.349609375,
      "learning_rate": 1.9776645942604487e-05,
      "loss": 0.5889,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1182
    },
    {
      "epoch": 0.16086483546369323,
      "grad_norm": 0.4453125,
      "learning_rate": 1.9775646998293333e-05,
      "loss": 0.8309,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1183
    },
    {
      "epoch": 0.16100081588251292,
      "grad_norm": 0.75390625,
      "learning_rate": 1.9774645850431828e-05,
      "loss": 0.7244,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1184
    },
    {
      "epoch": 0.1611367963013326,
      "grad_norm": 0.78515625,
      "learning_rate": 1.9773642499245645e-05,
      "loss": 0.6966,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1185
    },
    {
      "epoch": 0.1612727767201523,
      "grad_norm": 0.50390625,
      "learning_rate": 1.9772636944960947e-05,
      "loss": 0.8962,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1186
    },
    {
      "epoch": 0.161408757138972,
      "grad_norm": 0.40625,
      "learning_rate": 1.97716291878044e-05,
      "loss": 0.6491,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1187
    },
    {
      "epoch": 0.16154473755779167,
      "grad_norm": 0.44140625,
      "learning_rate": 1.977061922800317e-05,
      "loss": 0.6693,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1188
    },
    {
      "epoch": 0.16168071797661138,
      "grad_norm": 0.314453125,
      "learning_rate": 1.9769607065784906e-05,
      "loss": 0.4582,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1189
    },
    {
      "epoch": 0.16181669839543106,
      "grad_norm": 0.5390625,
      "learning_rate": 1.9768592701377765e-05,
      "loss": 0.416,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1190
    },
    {
      "epoch": 0.16195267881425074,
      "grad_norm": 0.4921875,
      "learning_rate": 1.97675761350104e-05,
      "loss": 0.6838,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1191
    },
    {
      "epoch": 0.16208865923307045,
      "grad_norm": 0.45703125,
      "learning_rate": 1.9766557366911952e-05,
      "loss": 0.6548,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1192
    },
    {
      "epoch": 0.16222463965189013,
      "grad_norm": 0.357421875,
      "learning_rate": 1.9765536397312067e-05,
      "loss": 0.6471,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1193
    },
    {
      "epoch": 0.1623606200707098,
      "grad_norm": 0.55859375,
      "learning_rate": 1.9764513226440888e-05,
      "loss": 0.5936,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1194
    },
    {
      "epoch": 0.16249660048952952,
      "grad_norm": 0.41015625,
      "learning_rate": 1.9763487854529042e-05,
      "loss": 0.7578,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1195
    },
    {
      "epoch": 0.1626325809083492,
      "grad_norm": 0.80078125,
      "learning_rate": 1.9762460281807665e-05,
      "loss": 0.759,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1196
    },
    {
      "epoch": 0.16276856132716888,
      "grad_norm": 0.3671875,
      "learning_rate": 1.9761430508508382e-05,
      "loss": 0.631,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1197
    },
    {
      "epoch": 0.1629045417459886,
      "grad_norm": 0.49609375,
      "learning_rate": 1.9760398534863324e-05,
      "loss": 0.7474,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1198
    },
    {
      "epoch": 0.16304052216480827,
      "grad_norm": 0.4609375,
      "learning_rate": 1.97593643611051e-05,
      "loss": 0.8389,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1199
    },
    {
      "epoch": 0.16317650258362795,
      "grad_norm": 0.400390625,
      "learning_rate": 1.9758327987466833e-05,
      "loss": 0.589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1200
    },
    {
      "epoch": 0.16331248300244763,
      "grad_norm": 0.44140625,
      "learning_rate": 1.975728941418213e-05,
      "loss": 0.7363,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1201
    },
    {
      "epoch": 0.16344846342126734,
      "grad_norm": 0.361328125,
      "learning_rate": 1.9756248641485102e-05,
      "loss": 0.5498,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1202
    },
    {
      "epoch": 0.16358444384008702,
      "grad_norm": 0.6953125,
      "learning_rate": 1.9755205669610347e-05,
      "loss": 0.8901,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1203
    },
    {
      "epoch": 0.1637204242589067,
      "grad_norm": 0.55859375,
      "learning_rate": 1.9754160498792964e-05,
      "loss": 0.9139,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1204
    },
    {
      "epoch": 0.1638564046777264,
      "grad_norm": 0.609375,
      "learning_rate": 1.9753113129268552e-05,
      "loss": 0.8415,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1205
    },
    {
      "epoch": 0.1639923850965461,
      "grad_norm": 0.5625,
      "learning_rate": 1.9752063561273198e-05,
      "loss": 0.7881,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1206
    },
    {
      "epoch": 0.16412836551536578,
      "grad_norm": 0.7421875,
      "learning_rate": 1.975101179504349e-05,
      "loss": 0.8063,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1207
    },
    {
      "epoch": 0.16426434593418549,
      "grad_norm": 0.4140625,
      "learning_rate": 1.9749957830816503e-05,
      "loss": 0.7756,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1208
    },
    {
      "epoch": 0.16440032635300517,
      "grad_norm": 0.3046875,
      "learning_rate": 1.9748901668829822e-05,
      "loss": 0.5106,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1209
    },
    {
      "epoch": 0.16453630677182485,
      "grad_norm": 0.275390625,
      "learning_rate": 1.974784330932151e-05,
      "loss": 0.5335,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1210
    },
    {
      "epoch": 0.16467228719064456,
      "grad_norm": 0.490234375,
      "learning_rate": 1.9746782752530138e-05,
      "loss": 0.8044,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1211
    },
    {
      "epoch": 0.16480826760946424,
      "grad_norm": 0.96484375,
      "learning_rate": 1.974571999869477e-05,
      "loss": 0.75,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1212
    },
    {
      "epoch": 0.16494424802828392,
      "grad_norm": 0.443359375,
      "learning_rate": 1.9744655048054962e-05,
      "loss": 0.6067,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1213
    },
    {
      "epoch": 0.16508022844710363,
      "grad_norm": 0.65234375,
      "learning_rate": 1.9743587900850767e-05,
      "loss": 0.7181,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1214
    },
    {
      "epoch": 0.1652162088659233,
      "grad_norm": 0.392578125,
      "learning_rate": 1.9742518557322735e-05,
      "loss": 0.744,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1215
    },
    {
      "epoch": 0.165352189284743,
      "grad_norm": 0.5078125,
      "learning_rate": 1.9741447017711905e-05,
      "loss": 0.7939,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1216
    },
    {
      "epoch": 0.1654881697035627,
      "grad_norm": 0.5390625,
      "learning_rate": 1.974037328225982e-05,
      "loss": 0.8475,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1217
    },
    {
      "epoch": 0.16562415012238238,
      "grad_norm": 1.0546875,
      "learning_rate": 1.9739297351208508e-05,
      "loss": 0.645,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1218
    },
    {
      "epoch": 0.16576013054120206,
      "grad_norm": 0.40234375,
      "learning_rate": 1.9738219224800503e-05,
      "loss": 0.7446,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1219
    },
    {
      "epoch": 0.16589611096002177,
      "grad_norm": 0.67578125,
      "learning_rate": 1.973713890327882e-05,
      "loss": 0.7642,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1220
    },
    {
      "epoch": 0.16603209137884145,
      "grad_norm": 0.65234375,
      "learning_rate": 1.9736056386886986e-05,
      "loss": 0.8324,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1221
    },
    {
      "epoch": 0.16616807179766113,
      "grad_norm": 0.392578125,
      "learning_rate": 1.9734971675869006e-05,
      "loss": 0.5917,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1222
    },
    {
      "epoch": 0.1663040522164808,
      "grad_norm": 0.515625,
      "learning_rate": 1.973388477046939e-05,
      "loss": 0.6859,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1223
    },
    {
      "epoch": 0.16644003263530052,
      "grad_norm": 0.345703125,
      "learning_rate": 1.973279567093314e-05,
      "loss": 0.5308,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1224
    },
    {
      "epoch": 0.1665760130541202,
      "grad_norm": 0.6171875,
      "learning_rate": 1.9731704377505746e-05,
      "loss": 0.8151,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1225
    },
    {
      "epoch": 0.16671199347293988,
      "grad_norm": 0.3671875,
      "learning_rate": 1.9730610890433212e-05,
      "loss": 0.757,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1226
    },
    {
      "epoch": 0.1668479738917596,
      "grad_norm": 0.38671875,
      "learning_rate": 1.9729515209962014e-05,
      "loss": 0.5674,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1227
    },
    {
      "epoch": 0.16698395431057927,
      "grad_norm": 0.54296875,
      "learning_rate": 1.9728417336339133e-05,
      "loss": 0.7894,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1228
    },
    {
      "epoch": 0.16711993472939896,
      "grad_norm": 0.361328125,
      "learning_rate": 1.972731726981205e-05,
      "loss": 0.6911,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1229
    },
    {
      "epoch": 0.16725591514821866,
      "grad_norm": 0.40234375,
      "learning_rate": 1.9726215010628717e-05,
      "loss": 0.7285,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1230
    },
    {
      "epoch": 0.16739189556703835,
      "grad_norm": 0.34765625,
      "learning_rate": 1.972511055903761e-05,
      "loss": 0.5993,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1231
    },
    {
      "epoch": 0.16752787598585803,
      "grad_norm": 0.5703125,
      "learning_rate": 1.972400391528769e-05,
      "loss": 0.9159,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1232
    },
    {
      "epoch": 0.16766385640467774,
      "grad_norm": 0.5703125,
      "learning_rate": 1.9722895079628395e-05,
      "loss": 0.4661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1233
    },
    {
      "epoch": 0.16779983682349742,
      "grad_norm": 0.41015625,
      "learning_rate": 1.9721784052309676e-05,
      "loss": 0.6217,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1234
    },
    {
      "epoch": 0.1679358172423171,
      "grad_norm": 0.388671875,
      "learning_rate": 1.9720670833581973e-05,
      "loss": 0.5798,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1235
    },
    {
      "epoch": 0.1680717976611368,
      "grad_norm": 0.310546875,
      "learning_rate": 1.971955542369622e-05,
      "loss": 0.5438,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1236
    },
    {
      "epoch": 0.1682077780799565,
      "grad_norm": 0.40234375,
      "learning_rate": 1.9718437822903844e-05,
      "loss": 0.6379,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1237
    },
    {
      "epoch": 0.16834375849877617,
      "grad_norm": 0.39453125,
      "learning_rate": 1.971731803145676e-05,
      "loss": 0.8447,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1238
    },
    {
      "epoch": 0.16847973891759588,
      "grad_norm": 0.5625,
      "learning_rate": 1.971619604960739e-05,
      "loss": 0.625,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1239
    },
    {
      "epoch": 0.16861571933641556,
      "grad_norm": 0.33984375,
      "learning_rate": 1.971507187760864e-05,
      "loss": 0.5498,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1240
    },
    {
      "epoch": 0.16875169975523524,
      "grad_norm": 0.359375,
      "learning_rate": 1.9713945515713907e-05,
      "loss": 0.6216,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1241
    },
    {
      "epoch": 0.16888768017405495,
      "grad_norm": 0.294921875,
      "learning_rate": 1.9712816964177098e-05,
      "loss": 0.4492,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1242
    },
    {
      "epoch": 0.16902366059287463,
      "grad_norm": 0.3515625,
      "learning_rate": 1.971168622325259e-05,
      "loss": 0.6309,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1243
    },
    {
      "epoch": 0.1691596410116943,
      "grad_norm": 0.349609375,
      "learning_rate": 1.9710553293195273e-05,
      "loss": 0.6245,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1244
    },
    {
      "epoch": 0.169295621430514,
      "grad_norm": 0.349609375,
      "learning_rate": 1.9709418174260523e-05,
      "loss": 0.61,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1245
    },
    {
      "epoch": 0.1694316018493337,
      "grad_norm": 0.361328125,
      "learning_rate": 1.9708280866704204e-05,
      "loss": 0.7505,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1246
    },
    {
      "epoch": 0.16956758226815338,
      "grad_norm": 0.427734375,
      "learning_rate": 1.970714137078269e-05,
      "loss": 0.6997,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1247
    },
    {
      "epoch": 0.16970356268697306,
      "grad_norm": 0.306640625,
      "learning_rate": 1.9705999686752827e-05,
      "loss": 0.5137,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1248
    },
    {
      "epoch": 0.16983954310579277,
      "grad_norm": 0.357421875,
      "learning_rate": 1.970485581487197e-05,
      "loss": 0.6011,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1249
    },
    {
      "epoch": 0.16997552352461245,
      "grad_norm": 0.453125,
      "learning_rate": 1.970370975539796e-05,
      "loss": 0.5189,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1250
    },
    {
      "epoch": 0.17011150394343214,
      "grad_norm": 0.361328125,
      "learning_rate": 1.970256150858913e-05,
      "loss": 0.5584,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1251
    },
    {
      "epoch": 0.17024748436225184,
      "grad_norm": 0.279296875,
      "learning_rate": 1.9701411074704312e-05,
      "loss": 0.5467,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1252
    },
    {
      "epoch": 0.17038346478107153,
      "grad_norm": 0.36328125,
      "learning_rate": 1.970025845400283e-05,
      "loss": 0.5731,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1253
    },
    {
      "epoch": 0.1705194451998912,
      "grad_norm": 0.31640625,
      "learning_rate": 1.9699103646744496e-05,
      "loss": 0.5856,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1254
    },
    {
      "epoch": 0.17065542561871092,
      "grad_norm": 0.400390625,
      "learning_rate": 1.9697946653189616e-05,
      "loss": 0.6045,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1255
    },
    {
      "epoch": 0.1707914060375306,
      "grad_norm": 0.333984375,
      "learning_rate": 1.9696787473598993e-05,
      "loss": 0.5635,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1256
    },
    {
      "epoch": 0.17092738645635028,
      "grad_norm": 0.36328125,
      "learning_rate": 1.969562610823392e-05,
      "loss": 0.6538,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1257
    },
    {
      "epoch": 0.17106336687517,
      "grad_norm": 0.40625,
      "learning_rate": 1.9694462557356185e-05,
      "loss": 0.5311,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1258
    },
    {
      "epoch": 0.17119934729398967,
      "grad_norm": 0.390625,
      "learning_rate": 1.969329682122806e-05,
      "loss": 0.6997,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1259
    },
    {
      "epoch": 0.17133532771280935,
      "grad_norm": 0.40234375,
      "learning_rate": 1.9692128900112325e-05,
      "loss": 0.7201,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1260
    },
    {
      "epoch": 0.17147130813162906,
      "grad_norm": 0.6171875,
      "learning_rate": 1.9690958794272237e-05,
      "loss": 0.7266,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1261
    },
    {
      "epoch": 0.17160728855044874,
      "grad_norm": 0.2265625,
      "learning_rate": 1.9689786503971553e-05,
      "loss": 0.4657,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1262
    },
    {
      "epoch": 0.17174326896926842,
      "grad_norm": 0.318359375,
      "learning_rate": 1.9688612029474527e-05,
      "loss": 0.6545,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1263
    },
    {
      "epoch": 0.17187924938808813,
      "grad_norm": 0.462890625,
      "learning_rate": 1.9687435371045893e-05,
      "loss": 0.6333,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1264
    },
    {
      "epoch": 0.1720152298069078,
      "grad_norm": 0.443359375,
      "learning_rate": 1.9686256528950887e-05,
      "loss": 0.6966,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1265
    },
    {
      "epoch": 0.1721512102257275,
      "grad_norm": 0.69140625,
      "learning_rate": 1.9685075503455232e-05,
      "loss": 0.6921,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1266
    },
    {
      "epoch": 0.17228719064454717,
      "grad_norm": 0.5625,
      "learning_rate": 1.968389229482515e-05,
      "loss": 0.5309,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1267
    },
    {
      "epoch": 0.17242317106336688,
      "grad_norm": 0.68359375,
      "learning_rate": 1.968270690332735e-05,
      "loss": 0.6935,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1268
    },
    {
      "epoch": 0.17255915148218656,
      "grad_norm": 0.328125,
      "learning_rate": 1.968151932922903e-05,
      "loss": 0.5221,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1269
    },
    {
      "epoch": 0.17269513190100624,
      "grad_norm": 0.91015625,
      "learning_rate": 1.968032957279789e-05,
      "loss": 0.6785,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1270
    },
    {
      "epoch": 0.17283111231982595,
      "grad_norm": 0.330078125,
      "learning_rate": 1.967913763430211e-05,
      "loss": 0.5513,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1271
    },
    {
      "epoch": 0.17296709273864563,
      "grad_norm": 0.388671875,
      "learning_rate": 1.9677943514010368e-05,
      "loss": 0.6324,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1272
    },
    {
      "epoch": 0.17310307315746531,
      "grad_norm": 0.33203125,
      "learning_rate": 1.967674721219184e-05,
      "loss": 0.4842,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1273
    },
    {
      "epoch": 0.17323905357628502,
      "grad_norm": 0.375,
      "learning_rate": 1.9675548729116176e-05,
      "loss": 0.7072,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1274
    },
    {
      "epoch": 0.1733750339951047,
      "grad_norm": 0.53515625,
      "learning_rate": 1.9674348065053533e-05,
      "loss": 0.755,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1275
    },
    {
      "epoch": 0.17351101441392439,
      "grad_norm": 0.369140625,
      "learning_rate": 1.967314522027456e-05,
      "loss": 0.5949,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1276
    },
    {
      "epoch": 0.1736469948327441,
      "grad_norm": 0.3359375,
      "learning_rate": 1.9671940195050393e-05,
      "loss": 0.6201,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1277
    },
    {
      "epoch": 0.17378297525156378,
      "grad_norm": 0.5078125,
      "learning_rate": 1.967073298965265e-05,
      "loss": 0.827,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1278
    },
    {
      "epoch": 0.17391895567038346,
      "grad_norm": 0.40234375,
      "learning_rate": 1.966952360435346e-05,
      "loss": 0.631,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1279
    },
    {
      "epoch": 0.17405493608920317,
      "grad_norm": 0.419921875,
      "learning_rate": 1.9668312039425425e-05,
      "loss": 0.7119,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1280
    },
    {
      "epoch": 0.17419091650802285,
      "grad_norm": 0.484375,
      "learning_rate": 1.9667098295141652e-05,
      "loss": 0.6484,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1281
    },
    {
      "epoch": 0.17432689692684253,
      "grad_norm": 0.50390625,
      "learning_rate": 1.9665882371775735e-05,
      "loss": 0.5741,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1282
    },
    {
      "epoch": 0.17446287734566224,
      "grad_norm": 0.50390625,
      "learning_rate": 1.9664664269601757e-05,
      "loss": 0.4888,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1283
    },
    {
      "epoch": 0.17459885776448192,
      "grad_norm": 0.625,
      "learning_rate": 1.966344398889429e-05,
      "loss": 0.5646,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1284
    },
    {
      "epoch": 0.1747348381833016,
      "grad_norm": 0.298828125,
      "learning_rate": 1.96622215299284e-05,
      "loss": 0.6134,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1285
    },
    {
      "epoch": 0.17487081860212128,
      "grad_norm": 0.69921875,
      "learning_rate": 1.9660996892979646e-05,
      "loss": 0.4948,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1286
    },
    {
      "epoch": 0.175006799020941,
      "grad_norm": 0.375,
      "learning_rate": 1.965977007832408e-05,
      "loss": 0.6069,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1287
    },
    {
      "epoch": 0.17514277943976067,
      "grad_norm": 0.51171875,
      "learning_rate": 1.965854108623823e-05,
      "loss": 0.6733,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1288
    },
    {
      "epoch": 0.17527875985858035,
      "grad_norm": 0.306640625,
      "learning_rate": 1.965730991699914e-05,
      "loss": 0.4141,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1289
    },
    {
      "epoch": 0.17541474027740006,
      "grad_norm": 0.625,
      "learning_rate": 1.965607657088432e-05,
      "loss": 0.6229,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1290
    },
    {
      "epoch": 0.17555072069621974,
      "grad_norm": 0.52734375,
      "learning_rate": 1.9654841048171786e-05,
      "loss": 0.7907,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1291
    },
    {
      "epoch": 0.17568670111503942,
      "grad_norm": 0.294921875,
      "learning_rate": 1.965360334914004e-05,
      "loss": 0.5654,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1292
    },
    {
      "epoch": 0.17582268153385913,
      "grad_norm": 0.29296875,
      "learning_rate": 1.9652363474068075e-05,
      "loss": 0.618,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1293
    },
    {
      "epoch": 0.1759586619526788,
      "grad_norm": 0.48828125,
      "learning_rate": 1.9651121423235367e-05,
      "loss": 0.9053,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1294
    },
    {
      "epoch": 0.1760946423714985,
      "grad_norm": 0.431640625,
      "learning_rate": 1.96498771969219e-05,
      "loss": 0.8005,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1295
    },
    {
      "epoch": 0.1762306227903182,
      "grad_norm": 0.6484375,
      "learning_rate": 1.964863079540813e-05,
      "loss": 0.6169,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1296
    },
    {
      "epoch": 0.17636660320913788,
      "grad_norm": 0.328125,
      "learning_rate": 1.9647382218975013e-05,
      "loss": 0.604,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1297
    },
    {
      "epoch": 0.17650258362795757,
      "grad_norm": 0.322265625,
      "learning_rate": 1.9646131467903995e-05,
      "loss": 0.5832,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1298
    },
    {
      "epoch": 0.17663856404677727,
      "grad_norm": 0.353515625,
      "learning_rate": 1.9644878542477012e-05,
      "loss": 0.743,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1299
    },
    {
      "epoch": 0.17677454446559696,
      "grad_norm": 0.6875,
      "learning_rate": 1.9643623442976482e-05,
      "loss": 0.7922,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1300
    },
    {
      "epoch": 0.17691052488441664,
      "grad_norm": 0.60546875,
      "learning_rate": 1.9642366169685328e-05,
      "loss": 0.7749,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1301
    },
    {
      "epoch": 0.17704650530323635,
      "grad_norm": 0.34765625,
      "learning_rate": 1.964110672288695e-05,
      "loss": 0.5609,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1302
    },
    {
      "epoch": 0.17718248572205603,
      "grad_norm": 0.72265625,
      "learning_rate": 1.9639845102865244e-05,
      "loss": 0.5804,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1303
    },
    {
      "epoch": 0.1773184661408757,
      "grad_norm": 0.314453125,
      "learning_rate": 1.963858130990459e-05,
      "loss": 0.5789,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1304
    },
    {
      "epoch": 0.17745444655969542,
      "grad_norm": 0.28515625,
      "learning_rate": 1.9637315344289874e-05,
      "loss": 0.4751,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1305
    },
    {
      "epoch": 0.1775904269785151,
      "grad_norm": 0.53125,
      "learning_rate": 1.9636047206306446e-05,
      "loss": 0.6872,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1306
    },
    {
      "epoch": 0.17772640739733478,
      "grad_norm": 0.490234375,
      "learning_rate": 1.9634776896240172e-05,
      "loss": 0.5967,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1307
    },
    {
      "epoch": 0.17786238781615446,
      "grad_norm": 0.412109375,
      "learning_rate": 1.963350441437739e-05,
      "loss": 0.5086,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1308
    },
    {
      "epoch": 0.17799836823497417,
      "grad_norm": 0.578125,
      "learning_rate": 1.963222976100493e-05,
      "loss": 0.8151,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1309
    },
    {
      "epoch": 0.17813434865379385,
      "grad_norm": 0.365234375,
      "learning_rate": 1.963095293641012e-05,
      "loss": 0.6279,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1310
    },
    {
      "epoch": 0.17827032907261353,
      "grad_norm": 0.259765625,
      "learning_rate": 1.9629673940880768e-05,
      "loss": 0.4538,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1311
    },
    {
      "epoch": 0.17840630949143324,
      "grad_norm": 0.33203125,
      "learning_rate": 1.9628392774705175e-05,
      "loss": 0.4754,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1312
    },
    {
      "epoch": 0.17854228991025292,
      "grad_norm": 0.48828125,
      "learning_rate": 1.9627109438172134e-05,
      "loss": 0.7018,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1313
    },
    {
      "epoch": 0.1786782703290726,
      "grad_norm": 0.38671875,
      "learning_rate": 1.9625823931570925e-05,
      "loss": 0.5936,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1314
    },
    {
      "epoch": 0.1788142507478923,
      "grad_norm": 0.357421875,
      "learning_rate": 1.9624536255191315e-05,
      "loss": 0.5033,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1315
    },
    {
      "epoch": 0.178950231166712,
      "grad_norm": 0.9375,
      "learning_rate": 1.962324640932356e-05,
      "loss": 0.9113,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1316
    },
    {
      "epoch": 0.17908621158553167,
      "grad_norm": 0.419921875,
      "learning_rate": 1.9621954394258414e-05,
      "loss": 0.562,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1317
    },
    {
      "epoch": 0.17922219200435138,
      "grad_norm": 0.458984375,
      "learning_rate": 1.9620660210287107e-05,
      "loss": 0.7033,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1318
    },
    {
      "epoch": 0.17935817242317106,
      "grad_norm": 0.396484375,
      "learning_rate": 1.9619363857701367e-05,
      "loss": 0.5534,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1319
    },
    {
      "epoch": 0.17949415284199075,
      "grad_norm": 0.2255859375,
      "learning_rate": 1.9618065336793407e-05,
      "loss": 0.3763,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1320
    },
    {
      "epoch": 0.17963013326081045,
      "grad_norm": 0.39453125,
      "learning_rate": 1.9616764647855924e-05,
      "loss": 0.5763,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1321
    },
    {
      "epoch": 0.17976611367963014,
      "grad_norm": 3.125,
      "learning_rate": 1.9615461791182118e-05,
      "loss": 0.9072,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1322
    },
    {
      "epoch": 0.17990209409844982,
      "grad_norm": 0.322265625,
      "learning_rate": 1.9614156767065667e-05,
      "loss": 0.6006,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1323
    },
    {
      "epoch": 0.18003807451726953,
      "grad_norm": 0.53515625,
      "learning_rate": 1.9612849575800733e-05,
      "loss": 0.918,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1324
    },
    {
      "epoch": 0.1801740549360892,
      "grad_norm": 0.34765625,
      "learning_rate": 1.9611540217681978e-05,
      "loss": 0.57,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1325
    },
    {
      "epoch": 0.1803100353549089,
      "grad_norm": 0.65625,
      "learning_rate": 1.9610228693004547e-05,
      "loss": 0.599,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1326
    },
    {
      "epoch": 0.1804460157737286,
      "grad_norm": 0.46484375,
      "learning_rate": 1.9608915002064073e-05,
      "loss": 0.721,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1327
    },
    {
      "epoch": 0.18058199619254828,
      "grad_norm": 0.201171875,
      "learning_rate": 1.9607599145156676e-05,
      "loss": 0.3973,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1328
    },
    {
      "epoch": 0.18071797661136796,
      "grad_norm": 0.412109375,
      "learning_rate": 1.9606281122578973e-05,
      "loss": 0.7218,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1329
    },
    {
      "epoch": 0.18085395703018764,
      "grad_norm": 0.4296875,
      "learning_rate": 1.9604960934628053e-05,
      "loss": 0.5866,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1330
    },
    {
      "epoch": 0.18098993744900735,
      "grad_norm": 0.29296875,
      "learning_rate": 1.9603638581601513e-05,
      "loss": 0.4051,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1331
    },
    {
      "epoch": 0.18112591786782703,
      "grad_norm": 0.330078125,
      "learning_rate": 1.960231406379742e-05,
      "loss": 0.5675,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1332
    },
    {
      "epoch": 0.1812618982866467,
      "grad_norm": 0.8203125,
      "learning_rate": 1.9600987381514334e-05,
      "loss": 0.7231,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1333
    },
    {
      "epoch": 0.18139787870546642,
      "grad_norm": 0.314453125,
      "learning_rate": 1.9599658535051312e-05,
      "loss": 0.4318,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1334
    },
    {
      "epoch": 0.1815338591242861,
      "grad_norm": 0.40625,
      "learning_rate": 1.959832752470789e-05,
      "loss": 0.77,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1335
    },
    {
      "epoch": 0.18166983954310578,
      "grad_norm": 0.388671875,
      "learning_rate": 1.9596994350784095e-05,
      "loss": 0.5801,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1336
    },
    {
      "epoch": 0.1818058199619255,
      "grad_norm": 0.55078125,
      "learning_rate": 1.9595659013580443e-05,
      "loss": 0.7863,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1337
    },
    {
      "epoch": 0.18194180038074517,
      "grad_norm": 0.349609375,
      "learning_rate": 1.959432151339793e-05,
      "loss": 0.5495,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1338
    },
    {
      "epoch": 0.18207778079956485,
      "grad_norm": 0.314453125,
      "learning_rate": 1.9592981850538046e-05,
      "loss": 0.6079,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1339
    },
    {
      "epoch": 0.18221376121838456,
      "grad_norm": 0.51171875,
      "learning_rate": 1.9591640025302765e-05,
      "loss": 0.5511,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1340
    },
    {
      "epoch": 0.18234974163720424,
      "grad_norm": 0.67578125,
      "learning_rate": 1.9590296037994557e-05,
      "loss": 0.6848,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1341
    },
    {
      "epoch": 0.18248572205602392,
      "grad_norm": 0.423828125,
      "learning_rate": 1.958894988891637e-05,
      "loss": 0.7144,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1342
    },
    {
      "epoch": 0.18262170247484363,
      "grad_norm": 0.345703125,
      "learning_rate": 1.958760157837164e-05,
      "loss": 0.5721,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1343
    },
    {
      "epoch": 0.18275768289366331,
      "grad_norm": 0.6953125,
      "learning_rate": 1.95862511066643e-05,
      "loss": 0.703,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1344
    },
    {
      "epoch": 0.182893663312483,
      "grad_norm": 0.27734375,
      "learning_rate": 1.9584898474098758e-05,
      "loss": 0.5343,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1345
    },
    {
      "epoch": 0.1830296437313027,
      "grad_norm": 0.80078125,
      "learning_rate": 1.9583543680979914e-05,
      "loss": 0.7145,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1346
    },
    {
      "epoch": 0.18316562415012239,
      "grad_norm": 0.291015625,
      "learning_rate": 1.9582186727613152e-05,
      "loss": 0.457,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1347
    },
    {
      "epoch": 0.18330160456894207,
      "grad_norm": 0.765625,
      "learning_rate": 1.9580827614304354e-05,
      "loss": 0.8818,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1348
    },
    {
      "epoch": 0.18343758498776178,
      "grad_norm": 0.296875,
      "learning_rate": 1.9579466341359873e-05,
      "loss": 0.5413,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1349
    },
    {
      "epoch": 0.18357356540658146,
      "grad_norm": 0.51171875,
      "learning_rate": 1.9578102909086558e-05,
      "loss": 0.813,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1350
    },
    {
      "epoch": 0.18370954582540114,
      "grad_norm": 0.796875,
      "learning_rate": 1.9576737317791745e-05,
      "loss": 0.7801,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1351
    },
    {
      "epoch": 0.18384552624422082,
      "grad_norm": 0.474609375,
      "learning_rate": 1.9575369567783258e-05,
      "loss": 0.7036,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1352
    },
    {
      "epoch": 0.18398150666304053,
      "grad_norm": 0.42578125,
      "learning_rate": 1.9573999659369396e-05,
      "loss": 0.611,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1353
    },
    {
      "epoch": 0.1841174870818602,
      "grad_norm": 0.357421875,
      "learning_rate": 1.9572627592858963e-05,
      "loss": 0.59,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1354
    },
    {
      "epoch": 0.1842534675006799,
      "grad_norm": 0.357421875,
      "learning_rate": 1.9571253368561232e-05,
      "loss": 0.6388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1355
    },
    {
      "epoch": 0.1843894479194996,
      "grad_norm": 0.357421875,
      "learning_rate": 1.9569876986785975e-05,
      "loss": 0.6089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1356
    },
    {
      "epoch": 0.18452542833831928,
      "grad_norm": 0.37109375,
      "learning_rate": 1.956849844784344e-05,
      "loss": 0.5965,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1357
    },
    {
      "epoch": 0.18466140875713896,
      "grad_norm": 0.34375,
      "learning_rate": 1.956711775204437e-05,
      "loss": 0.5506,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1358
    },
    {
      "epoch": 0.18479738917595867,
      "grad_norm": 0.333984375,
      "learning_rate": 1.9565734899699997e-05,
      "loss": 0.5034,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1359
    },
    {
      "epoch": 0.18493336959477835,
      "grad_norm": 0.251953125,
      "learning_rate": 1.9564349891122017e-05,
      "loss": 0.4676,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1360
    },
    {
      "epoch": 0.18506935001359803,
      "grad_norm": 0.5390625,
      "learning_rate": 1.956296272662264e-05,
      "loss": 0.549,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1361
    },
    {
      "epoch": 0.18520533043241774,
      "grad_norm": 0.39453125,
      "learning_rate": 1.956157340651455e-05,
      "loss": 0.7209,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1362
    },
    {
      "epoch": 0.18534131085123742,
      "grad_norm": 0.421875,
      "learning_rate": 1.9560181931110912e-05,
      "loss": 0.7223,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1363
    },
    {
      "epoch": 0.1854772912700571,
      "grad_norm": 0.8046875,
      "learning_rate": 1.955878830072538e-05,
      "loss": 0.7591,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1364
    },
    {
      "epoch": 0.1856132716888768,
      "grad_norm": 0.55078125,
      "learning_rate": 1.95573925156721e-05,
      "loss": 0.5838,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1365
    },
    {
      "epoch": 0.1857492521076965,
      "grad_norm": 0.322265625,
      "learning_rate": 1.9555994576265703e-05,
      "loss": 0.625,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1366
    },
    {
      "epoch": 0.18588523252651618,
      "grad_norm": 0.7265625,
      "learning_rate": 1.9554594482821293e-05,
      "loss": 1.0197,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1367
    },
    {
      "epoch": 0.18602121294533588,
      "grad_norm": 0.298828125,
      "learning_rate": 1.955319223565447e-05,
      "loss": 0.6595,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1368
    },
    {
      "epoch": 0.18615719336415557,
      "grad_norm": 0.48828125,
      "learning_rate": 1.9551787835081318e-05,
      "loss": 0.8224,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1369
    },
    {
      "epoch": 0.18629317378297525,
      "grad_norm": 0.26171875,
      "learning_rate": 1.955038128141841e-05,
      "loss": 0.488,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1370
    },
    {
      "epoch": 0.18642915420179493,
      "grad_norm": 0.408203125,
      "learning_rate": 1.9548972574982796e-05,
      "loss": 0.6195,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1371
    },
    {
      "epoch": 0.18656513462061464,
      "grad_norm": 0.578125,
      "learning_rate": 1.9547561716092013e-05,
      "loss": 0.9359,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1372
    },
    {
      "epoch": 0.18670111503943432,
      "grad_norm": 0.25,
      "learning_rate": 1.9546148705064097e-05,
      "loss": 0.4818,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1373
    },
    {
      "epoch": 0.186837095458254,
      "grad_norm": 0.40234375,
      "learning_rate": 1.9544733542217543e-05,
      "loss": 0.6382,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1374
    },
    {
      "epoch": 0.1869730758770737,
      "grad_norm": 0.421875,
      "learning_rate": 1.954331622787136e-05,
      "loss": 0.4618,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1375
    },
    {
      "epoch": 0.1871090562958934,
      "grad_norm": 0.400390625,
      "learning_rate": 1.9541896762345024e-05,
      "loss": 0.6795,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1376
    },
    {
      "epoch": 0.18724503671471307,
      "grad_norm": 0.33203125,
      "learning_rate": 1.9540475145958493e-05,
      "loss": 0.5553,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1377
    },
    {
      "epoch": 0.18738101713353278,
      "grad_norm": 0.6015625,
      "learning_rate": 1.953905137903222e-05,
      "loss": 0.6387,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1378
    },
    {
      "epoch": 0.18751699755235246,
      "grad_norm": 0.318359375,
      "learning_rate": 1.9537625461887147e-05,
      "loss": 0.4517,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1379
    },
    {
      "epoch": 0.18765297797117214,
      "grad_norm": 0.42578125,
      "learning_rate": 1.9536197394844682e-05,
      "loss": 0.7124,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1380
    },
    {
      "epoch": 0.18778895838999185,
      "grad_norm": 0.64453125,
      "learning_rate": 1.9534767178226738e-05,
      "loss": 0.9798,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1381
    },
    {
      "epoch": 0.18792493880881153,
      "grad_norm": 0.66796875,
      "learning_rate": 1.9533334812355694e-05,
      "loss": 0.5246,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1382
    },
    {
      "epoch": 0.1880609192276312,
      "grad_norm": 0.515625,
      "learning_rate": 1.9531900297554433e-05,
      "loss": 0.6313,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1383
    },
    {
      "epoch": 0.18819689964645092,
      "grad_norm": 0.5546875,
      "learning_rate": 1.9530463634146304e-05,
      "loss": 0.7884,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1384
    },
    {
      "epoch": 0.1883328800652706,
      "grad_norm": 0.58203125,
      "learning_rate": 1.9529024822455152e-05,
      "loss": 0.5721,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1385
    },
    {
      "epoch": 0.18846886048409028,
      "grad_norm": 0.373046875,
      "learning_rate": 1.9527583862805303e-05,
      "loss": 0.6118,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1386
    },
    {
      "epoch": 0.18860484090291,
      "grad_norm": 0.49609375,
      "learning_rate": 1.9526140755521566e-05,
      "loss": 0.9036,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1387
    },
    {
      "epoch": 0.18874082132172967,
      "grad_norm": 0.45703125,
      "learning_rate": 1.9524695500929236e-05,
      "loss": 0.6159,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1388
    },
    {
      "epoch": 0.18887680174054935,
      "grad_norm": 0.37890625,
      "learning_rate": 1.952324809935409e-05,
      "loss": 0.7119,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1389
    },
    {
      "epoch": 0.18901278215936906,
      "grad_norm": 0.408203125,
      "learning_rate": 1.952179855112239e-05,
      "loss": 0.7601,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1390
    },
    {
      "epoch": 0.18914876257818874,
      "grad_norm": 0.4921875,
      "learning_rate": 1.9520346856560885e-05,
      "loss": 0.7358,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1391
    },
    {
      "epoch": 0.18928474299700843,
      "grad_norm": 0.259765625,
      "learning_rate": 1.95188930159968e-05,
      "loss": 0.5049,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1392
    },
    {
      "epoch": 0.1894207234158281,
      "grad_norm": 0.6328125,
      "learning_rate": 1.9517437029757848e-05,
      "loss": 0.6234,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1393
    },
    {
      "epoch": 0.18955670383464782,
      "grad_norm": 0.390625,
      "learning_rate": 1.9515978898172233e-05,
      "loss": 0.596,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1394
    },
    {
      "epoch": 0.1896926842534675,
      "grad_norm": 0.5234375,
      "learning_rate": 1.951451862156863e-05,
      "loss": 0.6359,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1395
    },
    {
      "epoch": 0.18982866467228718,
      "grad_norm": 0.453125,
      "learning_rate": 1.95130562002762e-05,
      "loss": 0.7907,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1396
    },
    {
      "epoch": 0.1899646450911069,
      "grad_norm": 0.373046875,
      "learning_rate": 1.9511591634624603e-05,
      "loss": 0.6935,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1397
    },
    {
      "epoch": 0.19010062550992657,
      "grad_norm": 0.373046875,
      "learning_rate": 1.9510124924943956e-05,
      "loss": 0.5871,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1398
    },
    {
      "epoch": 0.19023660592874625,
      "grad_norm": 0.6484375,
      "learning_rate": 1.9508656071564883e-05,
      "loss": 0.7743,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1399
    },
    {
      "epoch": 0.19037258634756596,
      "grad_norm": 0.3828125,
      "learning_rate": 1.9507185074818475e-05,
      "loss": 0.7474,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1400
    },
    {
      "epoch": 0.19050856676638564,
      "grad_norm": 0.28515625,
      "learning_rate": 1.9505711935036315e-05,
      "loss": 0.4795,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1401
    },
    {
      "epoch": 0.19064454718520532,
      "grad_norm": 0.5234375,
      "learning_rate": 1.950423665255047e-05,
      "loss": 0.6595,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1402
    },
    {
      "epoch": 0.19078052760402503,
      "grad_norm": 0.470703125,
      "learning_rate": 1.950275922769348e-05,
      "loss": 0.6276,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1403
    },
    {
      "epoch": 0.1909165080228447,
      "grad_norm": 0.482421875,
      "learning_rate": 1.9501279660798383e-05,
      "loss": 0.7576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1404
    },
    {
      "epoch": 0.1910524884416644,
      "grad_norm": 0.61328125,
      "learning_rate": 1.9499797952198685e-05,
      "loss": 0.6211,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1405
    },
    {
      "epoch": 0.1911884688604841,
      "grad_norm": 0.435546875,
      "learning_rate": 1.9498314102228387e-05,
      "loss": 0.6748,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1406
    },
    {
      "epoch": 0.19132444927930378,
      "grad_norm": 0.40234375,
      "learning_rate": 1.9496828111221958e-05,
      "loss": 0.6948,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1407
    },
    {
      "epoch": 0.19146042969812346,
      "grad_norm": 0.34765625,
      "learning_rate": 1.9495339979514365e-05,
      "loss": 0.5109,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1408
    },
    {
      "epoch": 0.19159641011694317,
      "grad_norm": 0.515625,
      "learning_rate": 1.949384970744105e-05,
      "loss": 0.5425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1409
    },
    {
      "epoch": 0.19173239053576285,
      "grad_norm": 0.330078125,
      "learning_rate": 1.949235729533794e-05,
      "loss": 0.4445,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1410
    },
    {
      "epoch": 0.19186837095458253,
      "grad_norm": 0.67578125,
      "learning_rate": 1.9490862743541438e-05,
      "loss": 1.0132,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1411
    },
    {
      "epoch": 0.19200435137340224,
      "grad_norm": 0.40625,
      "learning_rate": 1.9489366052388443e-05,
      "loss": 0.6735,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1412
    },
    {
      "epoch": 0.19214033179222192,
      "grad_norm": 0.322265625,
      "learning_rate": 1.9487867222216315e-05,
      "loss": 0.5882,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1413
    },
    {
      "epoch": 0.1922763122110416,
      "grad_norm": 0.8359375,
      "learning_rate": 1.9486366253362924e-05,
      "loss": 0.7396,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1414
    },
    {
      "epoch": 0.1924122926298613,
      "grad_norm": 0.33203125,
      "learning_rate": 1.9484863146166593e-05,
      "loss": 0.6698,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1415
    },
    {
      "epoch": 0.192548273048681,
      "grad_norm": 0.482421875,
      "learning_rate": 1.9483357900966147e-05,
      "loss": 0.7293,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1416
    },
    {
      "epoch": 0.19268425346750068,
      "grad_norm": 0.6796875,
      "learning_rate": 1.9481850518100886e-05,
      "loss": 0.6774,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1417
    },
    {
      "epoch": 0.19282023388632036,
      "grad_norm": 0.421875,
      "learning_rate": 1.9480340997910593e-05,
      "loss": 0.6748,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1418
    },
    {
      "epoch": 0.19295621430514007,
      "grad_norm": 0.470703125,
      "learning_rate": 1.9478829340735527e-05,
      "loss": 0.916,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1419
    },
    {
      "epoch": 0.19309219472395975,
      "grad_norm": 0.46484375,
      "learning_rate": 1.9477315546916443e-05,
      "loss": 0.745,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1420
    },
    {
      "epoch": 0.19322817514277943,
      "grad_norm": 0.46484375,
      "learning_rate": 1.9475799616794567e-05,
      "loss": 0.8737,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1421
    },
    {
      "epoch": 0.19336415556159914,
      "grad_norm": 0.4375,
      "learning_rate": 1.94742815507116e-05,
      "loss": 0.7606,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1422
    },
    {
      "epoch": 0.19350013598041882,
      "grad_norm": 0.373046875,
      "learning_rate": 1.9472761349009742e-05,
      "loss": 0.7422,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1423
    },
    {
      "epoch": 0.1936361163992385,
      "grad_norm": 0.380859375,
      "learning_rate": 1.947123901203166e-05,
      "loss": 0.5425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1424
    },
    {
      "epoch": 0.1937720968180582,
      "grad_norm": 0.578125,
      "learning_rate": 1.946971454012051e-05,
      "loss": 0.8501,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1425
    },
    {
      "epoch": 0.1939080772368779,
      "grad_norm": 0.267578125,
      "learning_rate": 1.9468187933619923e-05,
      "loss": 0.4622,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1426
    },
    {
      "epoch": 0.19404405765569757,
      "grad_norm": 0.439453125,
      "learning_rate": 1.9466659192874016e-05,
      "loss": 0.6708,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1427
    },
    {
      "epoch": 0.19418003807451728,
      "grad_norm": 0.431640625,
      "learning_rate": 1.946512831822739e-05,
      "loss": 0.6343,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1428
    },
    {
      "epoch": 0.19431601849333696,
      "grad_norm": 0.423828125,
      "learning_rate": 1.9463595310025115e-05,
      "loss": 0.6794,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1429
    },
    {
      "epoch": 0.19445199891215664,
      "grad_norm": 0.462890625,
      "learning_rate": 1.946206016861276e-05,
      "loss": 0.8018,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1430
    },
    {
      "epoch": 0.19458797933097635,
      "grad_norm": 0.6796875,
      "learning_rate": 1.9460522894336354e-05,
      "loss": 0.7505,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1431
    },
    {
      "epoch": 0.19472395974979603,
      "grad_norm": 0.6171875,
      "learning_rate": 1.9458983487542425e-05,
      "loss": 0.7798,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1432
    },
    {
      "epoch": 0.19485994016861571,
      "grad_norm": 0.36328125,
      "learning_rate": 1.945744194857797e-05,
      "loss": 0.5347,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1433
    },
    {
      "epoch": 0.19499592058743542,
      "grad_norm": 0.640625,
      "learning_rate": 1.9455898277790478e-05,
      "loss": 0.7404,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1434
    },
    {
      "epoch": 0.1951319010062551,
      "grad_norm": 0.70703125,
      "learning_rate": 1.94543524755279e-05,
      "loss": 0.7957,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1435
    },
    {
      "epoch": 0.19526788142507479,
      "grad_norm": 0.5,
      "learning_rate": 1.9452804542138686e-05,
      "loss": 0.7381,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1436
    },
    {
      "epoch": 0.19540386184389447,
      "grad_norm": 0.455078125,
      "learning_rate": 1.9451254477971756e-05,
      "loss": 0.7557,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1437
    },
    {
      "epoch": 0.19553984226271418,
      "grad_norm": 0.4609375,
      "learning_rate": 1.9449702283376516e-05,
      "loss": 0.7677,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1438
    },
    {
      "epoch": 0.19567582268153386,
      "grad_norm": 0.6796875,
      "learning_rate": 1.944814795870285e-05,
      "loss": 0.6676,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1439
    },
    {
      "epoch": 0.19581180310035354,
      "grad_norm": 0.349609375,
      "learning_rate": 1.944659150430112e-05,
      "loss": 0.4697,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1440
    },
    {
      "epoch": 0.19594778351917325,
      "grad_norm": 0.6875,
      "learning_rate": 1.9445032920522168e-05,
      "loss": 0.8953,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1441
    },
    {
      "epoch": 0.19608376393799293,
      "grad_norm": 0.34765625,
      "learning_rate": 1.944347220771732e-05,
      "loss": 0.6493,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1442
    },
    {
      "epoch": 0.1962197443568126,
      "grad_norm": 0.365234375,
      "learning_rate": 1.9441909366238383e-05,
      "loss": 0.582,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1443
    },
    {
      "epoch": 0.19635572477563232,
      "grad_norm": 0.375,
      "learning_rate": 1.9440344396437635e-05,
      "loss": 0.6213,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1444
    },
    {
      "epoch": 0.196491705194452,
      "grad_norm": 0.376953125,
      "learning_rate": 1.943877729866784e-05,
      "loss": 0.6252,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1445
    },
    {
      "epoch": 0.19662768561327168,
      "grad_norm": 0.373046875,
      "learning_rate": 1.9437208073282247e-05,
      "loss": 0.6672,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1446
    },
    {
      "epoch": 0.1967636660320914,
      "grad_norm": 0.341796875,
      "learning_rate": 1.9435636720634573e-05,
      "loss": 0.6309,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1447
    },
    {
      "epoch": 0.19689964645091107,
      "grad_norm": 0.41015625,
      "learning_rate": 1.9434063241079017e-05,
      "loss": 0.6397,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1448
    },
    {
      "epoch": 0.19703562686973075,
      "grad_norm": 0.3671875,
      "learning_rate": 1.943248763497027e-05,
      "loss": 0.6021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1449
    },
    {
      "epoch": 0.19717160728855046,
      "grad_norm": 0.48828125,
      "learning_rate": 1.9430909902663485e-05,
      "loss": 0.7225,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1450
    },
    {
      "epoch": 0.19730758770737014,
      "grad_norm": 0.58203125,
      "learning_rate": 1.9429330044514305e-05,
      "loss": 0.6445,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1451
    },
    {
      "epoch": 0.19744356812618982,
      "grad_norm": 0.447265625,
      "learning_rate": 1.942774806087885e-05,
      "loss": 0.749,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1452
    },
    {
      "epoch": 0.19757954854500953,
      "grad_norm": 0.431640625,
      "learning_rate": 1.942616395211372e-05,
      "loss": 0.807,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1453
    },
    {
      "epoch": 0.1977155289638292,
      "grad_norm": 0.431640625,
      "learning_rate": 1.9424577718575987e-05,
      "loss": 0.6315,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1454
    },
    {
      "epoch": 0.1978515093826489,
      "grad_norm": 0.345703125,
      "learning_rate": 1.9422989360623207e-05,
      "loss": 0.5591,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1455
    },
    {
      "epoch": 0.19798748980146857,
      "grad_norm": 0.49609375,
      "learning_rate": 1.9421398878613423e-05,
      "loss": 0.7869,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1456
    },
    {
      "epoch": 0.19812347022028828,
      "grad_norm": 0.37890625,
      "learning_rate": 1.941980627290515e-05,
      "loss": 0.6336,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1457
    },
    {
      "epoch": 0.19825945063910796,
      "grad_norm": 0.369140625,
      "learning_rate": 1.941821154385737e-05,
      "loss": 0.6613,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1458
    },
    {
      "epoch": 0.19839543105792765,
      "grad_norm": 0.458984375,
      "learning_rate": 1.941661469182956e-05,
      "loss": 0.7394,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1459
    },
    {
      "epoch": 0.19853141147674735,
      "grad_norm": 0.294921875,
      "learning_rate": 1.9415015717181672e-05,
      "loss": 0.5921,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1460
    },
    {
      "epoch": 0.19866739189556704,
      "grad_norm": 0.400390625,
      "learning_rate": 1.941341462027413e-05,
      "loss": 0.6252,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1461
    },
    {
      "epoch": 0.19880337231438672,
      "grad_norm": 0.43359375,
      "learning_rate": 1.9411811401467846e-05,
      "loss": 0.6199,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1462
    },
    {
      "epoch": 0.19893935273320643,
      "grad_norm": 0.53515625,
      "learning_rate": 1.94102060611242e-05,
      "loss": 0.9323,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1463
    },
    {
      "epoch": 0.1990753331520261,
      "grad_norm": 0.375,
      "learning_rate": 1.9408598599605062e-05,
      "loss": 0.6133,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1464
    },
    {
      "epoch": 0.1992113135708458,
      "grad_norm": 0.60546875,
      "learning_rate": 1.940698901727277e-05,
      "loss": 0.728,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1465
    },
    {
      "epoch": 0.1993472939896655,
      "grad_norm": 0.384765625,
      "learning_rate": 1.9405377314490136e-05,
      "loss": 0.6769,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1466
    },
    {
      "epoch": 0.19948327440848518,
      "grad_norm": 0.7109375,
      "learning_rate": 1.9403763491620468e-05,
      "loss": 0.5602,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1467
    },
    {
      "epoch": 0.19961925482730486,
      "grad_norm": 0.423828125,
      "learning_rate": 1.940214754902754e-05,
      "loss": 0.7778,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1468
    },
    {
      "epoch": 0.19975523524612457,
      "grad_norm": 0.45703125,
      "learning_rate": 1.94005294870756e-05,
      "loss": 0.7255,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1469
    },
    {
      "epoch": 0.19989121566494425,
      "grad_norm": 0.84375,
      "learning_rate": 1.9398909306129386e-05,
      "loss": 0.7798,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1470
    },
    {
      "epoch": 0.20002719608376393,
      "grad_norm": 0.62109375,
      "learning_rate": 1.9397287006554096e-05,
      "loss": 0.8716,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1471
    },
    {
      "epoch": 0.20016317650258364,
      "grad_norm": 0.8515625,
      "learning_rate": 1.9395662588715428e-05,
      "loss": 0.793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1472
    },
    {
      "epoch": 0.20029915692140332,
      "grad_norm": 0.5234375,
      "learning_rate": 1.939403605297954e-05,
      "loss": 0.7598,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1473
    },
    {
      "epoch": 0.200435137340223,
      "grad_norm": 0.53515625,
      "learning_rate": 1.9392407399713072e-05,
      "loss": 0.6774,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1474
    },
    {
      "epoch": 0.2005711177590427,
      "grad_norm": 0.61328125,
      "learning_rate": 1.9390776629283143e-05,
      "loss": 0.5526,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1475
    },
    {
      "epoch": 0.2007070981778624,
      "grad_norm": 0.2392578125,
      "learning_rate": 1.938914374205735e-05,
      "loss": 0.3314,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1476
    },
    {
      "epoch": 0.20084307859668207,
      "grad_norm": 0.416015625,
      "learning_rate": 1.938750873840377e-05,
      "loss": 0.7155,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1477
    },
    {
      "epoch": 0.20097905901550175,
      "grad_norm": 0.59375,
      "learning_rate": 1.938587161869094e-05,
      "loss": 0.5868,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1478
    },
    {
      "epoch": 0.20111503943432146,
      "grad_norm": 0.279296875,
      "learning_rate": 1.9384232383287898e-05,
      "loss": 0.5439,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1479
    },
    {
      "epoch": 0.20125101985314114,
      "grad_norm": 0.34375,
      "learning_rate": 1.9382591032564146e-05,
      "loss": 0.5,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1480
    },
    {
      "epoch": 0.20138700027196083,
      "grad_norm": 0.6328125,
      "learning_rate": 1.9380947566889663e-05,
      "loss": 0.6787,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1481
    },
    {
      "epoch": 0.20152298069078053,
      "grad_norm": 0.4921875,
      "learning_rate": 1.9379301986634906e-05,
      "loss": 0.7417,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1482
    },
    {
      "epoch": 0.20165896110960022,
      "grad_norm": 0.50390625,
      "learning_rate": 1.9377654292170808e-05,
      "loss": 0.791,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1483
    },
    {
      "epoch": 0.2017949415284199,
      "grad_norm": 0.462890625,
      "learning_rate": 1.9376004483868777e-05,
      "loss": 0.6834,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1484
    },
    {
      "epoch": 0.2019309219472396,
      "grad_norm": 0.318359375,
      "learning_rate": 1.937435256210071e-05,
      "loss": 0.5902,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1485
    },
    {
      "epoch": 0.2020669023660593,
      "grad_norm": 0.6484375,
      "learning_rate": 1.937269852723896e-05,
      "loss": 0.5557,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1486
    },
    {
      "epoch": 0.20220288278487897,
      "grad_norm": 0.2578125,
      "learning_rate": 1.9371042379656372e-05,
      "loss": 0.443,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1487
    },
    {
      "epoch": 0.20233886320369868,
      "grad_norm": 0.31640625,
      "learning_rate": 1.936938411972626e-05,
      "loss": 0.547,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1488
    },
    {
      "epoch": 0.20247484362251836,
      "grad_norm": 0.486328125,
      "learning_rate": 1.9367723747822418e-05,
      "loss": 0.7542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1489
    },
    {
      "epoch": 0.20261082404133804,
      "grad_norm": 0.310546875,
      "learning_rate": 1.9366061264319112e-05,
      "loss": 0.5721,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1490
    },
    {
      "epoch": 0.20274680446015775,
      "grad_norm": 0.384765625,
      "learning_rate": 1.9364396669591084e-05,
      "loss": 0.7017,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1491
    },
    {
      "epoch": 0.20288278487897743,
      "grad_norm": 0.32421875,
      "learning_rate": 1.936272996401356e-05,
      "loss": 0.6418,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1492
    },
    {
      "epoch": 0.2030187652977971,
      "grad_norm": 0.57421875,
      "learning_rate": 1.9361061147962233e-05,
      "loss": 0.6605,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1493
    },
    {
      "epoch": 0.20315474571661682,
      "grad_norm": 0.34375,
      "learning_rate": 1.9359390221813272e-05,
      "loss": 0.6934,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1494
    },
    {
      "epoch": 0.2032907261354365,
      "grad_norm": 0.78125,
      "learning_rate": 1.9357717185943328e-05,
      "loss": 0.7106,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1495
    },
    {
      "epoch": 0.20342670655425618,
      "grad_norm": 0.49609375,
      "learning_rate": 1.9356042040729522e-05,
      "loss": 0.779,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1496
    },
    {
      "epoch": 0.2035626869730759,
      "grad_norm": 0.76953125,
      "learning_rate": 1.9354364786549452e-05,
      "loss": 0.8325,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1497
    },
    {
      "epoch": 0.20369866739189557,
      "grad_norm": 0.369140625,
      "learning_rate": 1.9352685423781193e-05,
      "loss": 0.6572,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1498
    },
    {
      "epoch": 0.20383464781071525,
      "grad_norm": 0.353515625,
      "learning_rate": 1.935100395280329e-05,
      "loss": 0.5475,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1499
    },
    {
      "epoch": 0.20397062822953493,
      "grad_norm": 0.435546875,
      "learning_rate": 1.934932037399477e-05,
      "loss": 0.641,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1500
    },
    {
      "epoch": 0.20410660864835464,
      "grad_norm": 0.66796875,
      "learning_rate": 1.9347634687735134e-05,
      "loss": 0.8372,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1501
    },
    {
      "epoch": 0.20424258906717432,
      "grad_norm": 0.4609375,
      "learning_rate": 1.9345946894404356e-05,
      "loss": 0.4676,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1502
    },
    {
      "epoch": 0.204378569485994,
      "grad_norm": 0.328125,
      "learning_rate": 1.934425699438288e-05,
      "loss": 0.6279,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1503
    },
    {
      "epoch": 0.2045145499048137,
      "grad_norm": 0.490234375,
      "learning_rate": 1.9342564988051636e-05,
      "loss": 0.624,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1504
    },
    {
      "epoch": 0.2046505303236334,
      "grad_norm": 0.6171875,
      "learning_rate": 1.934087087579202e-05,
      "loss": 0.7944,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1505
    },
    {
      "epoch": 0.20478651074245308,
      "grad_norm": 0.388671875,
      "learning_rate": 1.9339174657985905e-05,
      "loss": 0.6396,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1506
    },
    {
      "epoch": 0.20492249116127279,
      "grad_norm": 0.61328125,
      "learning_rate": 1.9337476335015646e-05,
      "loss": 0.8415,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1507
    },
    {
      "epoch": 0.20505847158009247,
      "grad_norm": 0.44921875,
      "learning_rate": 1.9335775907264054e-05,
      "loss": 0.6351,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1508
    },
    {
      "epoch": 0.20519445199891215,
      "grad_norm": 0.443359375,
      "learning_rate": 1.9334073375114437e-05,
      "loss": 0.4246,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1509
    },
    {
      "epoch": 0.20533043241773186,
      "grad_norm": 0.546875,
      "learning_rate": 1.933236873895056e-05,
      "loss": 0.6182,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1510
    },
    {
      "epoch": 0.20546641283655154,
      "grad_norm": 0.328125,
      "learning_rate": 1.933066199915667e-05,
      "loss": 0.4964,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1511
    },
    {
      "epoch": 0.20560239325537122,
      "grad_norm": 0.3671875,
      "learning_rate": 1.9328953156117492e-05,
      "loss": 0.5822,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1512
    },
    {
      "epoch": 0.20573837367419093,
      "grad_norm": 0.81640625,
      "learning_rate": 1.932724221021821e-05,
      "loss": 0.8841,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1513
    },
    {
      "epoch": 0.2058743540930106,
      "grad_norm": 0.341796875,
      "learning_rate": 1.9325529161844502e-05,
      "loss": 0.5324,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1514
    },
    {
      "epoch": 0.2060103345118303,
      "grad_norm": 0.50390625,
      "learning_rate": 1.9323814011382508e-05,
      "loss": 0.7077,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1515
    },
    {
      "epoch": 0.20614631493065,
      "grad_norm": 0.31640625,
      "learning_rate": 1.9322096759218838e-05,
      "loss": 0.5338,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1516
    },
    {
      "epoch": 0.20628229534946968,
      "grad_norm": 0.34765625,
      "learning_rate": 1.932037740574059e-05,
      "loss": 0.4665,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1517
    },
    {
      "epoch": 0.20641827576828936,
      "grad_norm": 0.357421875,
      "learning_rate": 1.931865595133532e-05,
      "loss": 0.6693,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1518
    },
    {
      "epoch": 0.20655425618710907,
      "grad_norm": 0.51953125,
      "learning_rate": 1.931693239639107e-05,
      "loss": 0.6102,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1519
    },
    {
      "epoch": 0.20669023660592875,
      "grad_norm": 0.498046875,
      "learning_rate": 1.9315206741296345e-05,
      "loss": 0.6219,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1520
    },
    {
      "epoch": 0.20682621702474843,
      "grad_norm": 0.361328125,
      "learning_rate": 1.9313478986440134e-05,
      "loss": 0.6854,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1521
    },
    {
      "epoch": 0.2069621974435681,
      "grad_norm": 0.32421875,
      "learning_rate": 1.9311749132211893e-05,
      "loss": 0.4556,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1522
    },
    {
      "epoch": 0.20709817786238782,
      "grad_norm": 0.59765625,
      "learning_rate": 1.931001717900155e-05,
      "loss": 1.0825,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1523
    },
    {
      "epoch": 0.2072341582812075,
      "grad_norm": 0.41796875,
      "learning_rate": 1.930828312719951e-05,
      "loss": 0.7287,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1524
    },
    {
      "epoch": 0.20737013870002718,
      "grad_norm": 0.298828125,
      "learning_rate": 1.930654697719665e-05,
      "loss": 0.4769,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1525
    },
    {
      "epoch": 0.2075061191188469,
      "grad_norm": 0.98828125,
      "learning_rate": 1.930480872938432e-05,
      "loss": 0.7689,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1526
    },
    {
      "epoch": 0.20764209953766657,
      "grad_norm": 0.3046875,
      "learning_rate": 1.9303068384154338e-05,
      "loss": 0.5952,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1527
    },
    {
      "epoch": 0.20777807995648626,
      "grad_norm": 0.609375,
      "learning_rate": 1.9301325941899004e-05,
      "loss": 0.9837,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1528
    },
    {
      "epoch": 0.20791406037530596,
      "grad_norm": 0.404296875,
      "learning_rate": 1.9299581403011082e-05,
      "loss": 0.681,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1529
    },
    {
      "epoch": 0.20805004079412565,
      "grad_norm": 0.37890625,
      "learning_rate": 1.929783476788382e-05,
      "loss": 0.6592,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1530
    },
    {
      "epoch": 0.20818602121294533,
      "grad_norm": 0.7265625,
      "learning_rate": 1.929608603691092e-05,
      "loss": 0.7318,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1531
    },
    {
      "epoch": 0.20832200163176504,
      "grad_norm": 0.6171875,
      "learning_rate": 1.929433521048657e-05,
      "loss": 0.5441,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1532
    },
    {
      "epoch": 0.20845798205058472,
      "grad_norm": 0.58984375,
      "learning_rate": 1.9292582289005436e-05,
      "loss": 0.6839,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1533
    },
    {
      "epoch": 0.2085939624694044,
      "grad_norm": 0.765625,
      "learning_rate": 1.929082727286264e-05,
      "loss": 0.7209,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1534
    },
    {
      "epoch": 0.2087299428882241,
      "grad_norm": 0.9296875,
      "learning_rate": 1.9289070162453787e-05,
      "loss": 0.8143,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1535
    },
    {
      "epoch": 0.2088659233070438,
      "grad_norm": 0.396484375,
      "learning_rate": 1.928731095817495e-05,
      "loss": 0.4707,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1536
    },
    {
      "epoch": 0.20900190372586347,
      "grad_norm": 0.482421875,
      "learning_rate": 1.928554966042268e-05,
      "loss": 0.5391,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1537
    },
    {
      "epoch": 0.20913788414468318,
      "grad_norm": 0.408203125,
      "learning_rate": 1.9283786269593984e-05,
      "loss": 0.6688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1538
    },
    {
      "epoch": 0.20927386456350286,
      "grad_norm": 0.6640625,
      "learning_rate": 1.928202078608636e-05,
      "loss": 0.8105,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1539
    },
    {
      "epoch": 0.20940984498232254,
      "grad_norm": 0.34375,
      "learning_rate": 1.928025321029777e-05,
      "loss": 0.5859,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1540
    },
    {
      "epoch": 0.20954582540114222,
      "grad_norm": 0.353515625,
      "learning_rate": 1.9278483542626642e-05,
      "loss": 0.4808,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1541
    },
    {
      "epoch": 0.20968180581996193,
      "grad_norm": 0.5859375,
      "learning_rate": 1.9276711783471888e-05,
      "loss": 0.8424,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1542
    },
    {
      "epoch": 0.2098177862387816,
      "grad_norm": 0.384765625,
      "learning_rate": 1.9274937933232877e-05,
      "loss": 0.6576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1543
    },
    {
      "epoch": 0.2099537666576013,
      "grad_norm": 0.37109375,
      "learning_rate": 1.9273161992309462e-05,
      "loss": 0.6745,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1544
    },
    {
      "epoch": 0.210089747076421,
      "grad_norm": 0.49609375,
      "learning_rate": 1.927138396110196e-05,
      "loss": 0.6576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1545
    },
    {
      "epoch": 0.21022572749524068,
      "grad_norm": 0.66796875,
      "learning_rate": 1.9269603840011157e-05,
      "loss": 0.6141,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1546
    },
    {
      "epoch": 0.21036170791406036,
      "grad_norm": 0.44140625,
      "learning_rate": 1.926782162943832e-05,
      "loss": 0.5529,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1547
    },
    {
      "epoch": 0.21049768833288007,
      "grad_norm": 0.240234375,
      "learning_rate": 1.9266037329785183e-05,
      "loss": 0.4284,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1548
    },
    {
      "epoch": 0.21063366875169975,
      "grad_norm": 0.376953125,
      "learning_rate": 1.9264250941453937e-05,
      "loss": 0.6831,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1549
    },
    {
      "epoch": 0.21076964917051944,
      "grad_norm": 0.375,
      "learning_rate": 1.9262462464847263e-05,
      "loss": 0.6234,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1550
    },
    {
      "epoch": 0.21090562958933914,
      "grad_norm": 0.71484375,
      "learning_rate": 1.9260671900368312e-05,
      "loss": 0.7607,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1551
    },
    {
      "epoch": 0.21104161000815883,
      "grad_norm": 0.625,
      "learning_rate": 1.925887924842069e-05,
      "loss": 0.5199,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1552
    },
    {
      "epoch": 0.2111775904269785,
      "grad_norm": 0.3984375,
      "learning_rate": 1.9257084509408486e-05,
      "loss": 0.6774,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1553
    },
    {
      "epoch": 0.21131357084579822,
      "grad_norm": 0.41015625,
      "learning_rate": 1.9255287683736257e-05,
      "loss": 0.6619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1554
    },
    {
      "epoch": 0.2114495512646179,
      "grad_norm": 0.5078125,
      "learning_rate": 1.9253488771809024e-05,
      "loss": 0.7539,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1555
    },
    {
      "epoch": 0.21158553168343758,
      "grad_norm": 0.27734375,
      "learning_rate": 1.9251687774032292e-05,
      "loss": 0.4176,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1556
    },
    {
      "epoch": 0.2117215121022573,
      "grad_norm": 0.546875,
      "learning_rate": 1.924988469081202e-05,
      "loss": 0.7358,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1557
    },
    {
      "epoch": 0.21185749252107697,
      "grad_norm": 0.478515625,
      "learning_rate": 1.924807952255465e-05,
      "loss": 0.8039,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1558
    },
    {
      "epoch": 0.21199347293989665,
      "grad_norm": 0.69140625,
      "learning_rate": 1.9246272269667093e-05,
      "loss": 0.7581,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1559
    },
    {
      "epoch": 0.21212945335871636,
      "grad_norm": 0.416015625,
      "learning_rate": 1.9244462932556717e-05,
      "loss": 0.7186,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1560
    },
    {
      "epoch": 0.21226543377753604,
      "grad_norm": 0.3984375,
      "learning_rate": 1.924265151163137e-05,
      "loss": 0.7222,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1561
    },
    {
      "epoch": 0.21240141419635572,
      "grad_norm": 1.796875,
      "learning_rate": 1.9240838007299375e-05,
      "loss": 0.8831,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1562
    },
    {
      "epoch": 0.2125373946151754,
      "grad_norm": 0.330078125,
      "learning_rate": 1.923902241996951e-05,
      "loss": 0.5946,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1563
    },
    {
      "epoch": 0.2126733750339951,
      "grad_norm": 0.51953125,
      "learning_rate": 1.9237204750051036e-05,
      "loss": 0.7863,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1564
    },
    {
      "epoch": 0.2128093554528148,
      "grad_norm": 0.353515625,
      "learning_rate": 1.923538499795368e-05,
      "loss": 0.6263,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1565
    },
    {
      "epoch": 0.21294533587163447,
      "grad_norm": 0.48828125,
      "learning_rate": 1.923356316408763e-05,
      "loss": 0.6442,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1566
    },
    {
      "epoch": 0.21308131629045418,
      "grad_norm": 0.451171875,
      "learning_rate": 1.9231739248863553e-05,
      "loss": 0.5471,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1567
    },
    {
      "epoch": 0.21321729670927386,
      "grad_norm": 0.41796875,
      "learning_rate": 1.922991325269258e-05,
      "loss": 0.7946,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1568
    },
    {
      "epoch": 0.21335327712809354,
      "grad_norm": 0.8203125,
      "learning_rate": 1.9228085175986316e-05,
      "loss": 0.7528,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1569
    },
    {
      "epoch": 0.21348925754691325,
      "grad_norm": 0.349609375,
      "learning_rate": 1.922625501915683e-05,
      "loss": 0.6296,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1570
    },
    {
      "epoch": 0.21362523796573293,
      "grad_norm": 0.369140625,
      "learning_rate": 1.922442278261666e-05,
      "loss": 0.6392,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1571
    },
    {
      "epoch": 0.21376121838455261,
      "grad_norm": 0.66015625,
      "learning_rate": 1.9222588466778815e-05,
      "loss": 0.6735,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1572
    },
    {
      "epoch": 0.21389719880337232,
      "grad_norm": 0.287109375,
      "learning_rate": 1.9220752072056777e-05,
      "loss": 0.4683,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1573
    },
    {
      "epoch": 0.214033179222192,
      "grad_norm": 0.41796875,
      "learning_rate": 1.921891359886449e-05,
      "loss": 0.6978,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1574
    },
    {
      "epoch": 0.21416915964101169,
      "grad_norm": 0.3515625,
      "learning_rate": 1.9217073047616363e-05,
      "loss": 0.4738,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1575
    },
    {
      "epoch": 0.2143051400598314,
      "grad_norm": 0.57421875,
      "learning_rate": 1.9215230418727282e-05,
      "loss": 0.832,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1576
    },
    {
      "epoch": 0.21444112047865108,
      "grad_norm": 0.34765625,
      "learning_rate": 1.9213385712612597e-05,
      "loss": 0.598,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1577
    },
    {
      "epoch": 0.21457710089747076,
      "grad_norm": 0.30078125,
      "learning_rate": 1.921153892968813e-05,
      "loss": 0.6291,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1578
    },
    {
      "epoch": 0.21471308131629047,
      "grad_norm": 0.71484375,
      "learning_rate": 1.9209690070370172e-05,
      "loss": 0.644,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1579
    },
    {
      "epoch": 0.21484906173511015,
      "grad_norm": 0.326171875,
      "learning_rate": 1.920783913507547e-05,
      "loss": 0.6525,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1580
    },
    {
      "epoch": 0.21498504215392983,
      "grad_norm": 0.337890625,
      "learning_rate": 1.920598612422125e-05,
      "loss": 0.5793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1581
    },
    {
      "epoch": 0.21512102257274954,
      "grad_norm": 0.48046875,
      "learning_rate": 1.9204131038225207e-05,
      "loss": 0.6004,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1582
    },
    {
      "epoch": 0.21525700299156922,
      "grad_norm": 0.578125,
      "learning_rate": 1.9202273877505496e-05,
      "loss": 0.8721,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1583
    },
    {
      "epoch": 0.2153929834103889,
      "grad_norm": 0.294921875,
      "learning_rate": 1.9200414642480747e-05,
      "loss": 0.6379,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1584
    },
    {
      "epoch": 0.21552896382920858,
      "grad_norm": 0.458984375,
      "learning_rate": 1.9198553333570053e-05,
      "loss": 0.7025,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1585
    },
    {
      "epoch": 0.2156649442480283,
      "grad_norm": 0.59375,
      "learning_rate": 1.919668995119297e-05,
      "loss": 0.4209,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1586
    },
    {
      "epoch": 0.21580092466684797,
      "grad_norm": 0.36328125,
      "learning_rate": 1.9194824495769536e-05,
      "loss": 0.6421,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1587
    },
    {
      "epoch": 0.21593690508566765,
      "grad_norm": 0.29296875,
      "learning_rate": 1.9192956967720245e-05,
      "loss": 0.578,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1588
    },
    {
      "epoch": 0.21607288550448736,
      "grad_norm": 0.412109375,
      "learning_rate": 1.919108736746606e-05,
      "loss": 0.8073,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1589
    },
    {
      "epoch": 0.21620886592330704,
      "grad_norm": 0.44921875,
      "learning_rate": 1.9189215695428412e-05,
      "loss": 0.7202,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1590
    },
    {
      "epoch": 0.21634484634212672,
      "grad_norm": 0.2021484375,
      "learning_rate": 1.9187341952029194e-05,
      "loss": 0.2918,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1591
    },
    {
      "epoch": 0.21648082676094643,
      "grad_norm": 0.41796875,
      "learning_rate": 1.9185466137690777e-05,
      "loss": 0.6514,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1592
    },
    {
      "epoch": 0.2166168071797661,
      "grad_norm": 0.34375,
      "learning_rate": 1.918358825283599e-05,
      "loss": 0.5601,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1593
    },
    {
      "epoch": 0.2167527875985858,
      "grad_norm": 1.2578125,
      "learning_rate": 1.9181708297888133e-05,
      "loss": 0.7471,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1594
    },
    {
      "epoch": 0.2168887680174055,
      "grad_norm": 0.3828125,
      "learning_rate": 1.917982627327097e-05,
      "loss": 0.6268,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1595
    },
    {
      "epoch": 0.21702474843622518,
      "grad_norm": 0.4453125,
      "learning_rate": 1.9177942179408727e-05,
      "loss": 0.7808,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1596
    },
    {
      "epoch": 0.21716072885504487,
      "grad_norm": 0.443359375,
      "learning_rate": 1.9176056016726113e-05,
      "loss": 0.7178,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1597
    },
    {
      "epoch": 0.21729670927386457,
      "grad_norm": 0.443359375,
      "learning_rate": 1.917416778564828e-05,
      "loss": 0.7728,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1598
    },
    {
      "epoch": 0.21743268969268426,
      "grad_norm": 0.439453125,
      "learning_rate": 1.917227748660087e-05,
      "loss": 0.6914,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1599
    },
    {
      "epoch": 0.21756867011150394,
      "grad_norm": 0.6328125,
      "learning_rate": 1.917038512000997e-05,
      "loss": 0.5109,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1600
    },
    {
      "epoch": 0.21770465053032365,
      "grad_norm": 0.365234375,
      "learning_rate": 1.9168490686302144e-05,
      "loss": 0.6484,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1601
    },
    {
      "epoch": 0.21784063094914333,
      "grad_norm": 0.26953125,
      "learning_rate": 1.9166594185904427e-05,
      "loss": 0.5177,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1602
    },
    {
      "epoch": 0.217976611367963,
      "grad_norm": 0.404296875,
      "learning_rate": 1.9164695619244303e-05,
      "loss": 0.6105,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1603
    },
    {
      "epoch": 0.21811259178678272,
      "grad_norm": 0.3984375,
      "learning_rate": 1.9162794986749744e-05,
      "loss": 0.6982,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1604
    },
    {
      "epoch": 0.2182485722056024,
      "grad_norm": 0.671875,
      "learning_rate": 1.9160892288849168e-05,
      "loss": 0.4574,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1605
    },
    {
      "epoch": 0.21838455262442208,
      "grad_norm": 0.3671875,
      "learning_rate": 1.915898752597147e-05,
      "loss": 0.6748,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1606
    },
    {
      "epoch": 0.21852053304324176,
      "grad_norm": 0.390625,
      "learning_rate": 1.9157080698546e-05,
      "loss": 0.5054,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1607
    },
    {
      "epoch": 0.21865651346206147,
      "grad_norm": 0.33203125,
      "learning_rate": 1.9155171807002587e-05,
      "loss": 0.63,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1608
    },
    {
      "epoch": 0.21879249388088115,
      "grad_norm": 0.275390625,
      "learning_rate": 1.915326085177152e-05,
      "loss": 0.502,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1609
    },
    {
      "epoch": 0.21892847429970083,
      "grad_norm": 0.328125,
      "learning_rate": 1.9151347833283545e-05,
      "loss": 0.5355,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1610
    },
    {
      "epoch": 0.21906445471852054,
      "grad_norm": 0.5625,
      "learning_rate": 1.9149432751969885e-05,
      "loss": 0.7925,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1611
    },
    {
      "epoch": 0.21920043513734022,
      "grad_norm": 0.388671875,
      "learning_rate": 1.9147515608262224e-05,
      "loss": 0.6997,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1612
    },
    {
      "epoch": 0.2193364155561599,
      "grad_norm": 0.375,
      "learning_rate": 1.9145596402592704e-05,
      "loss": 0.6453,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1613
    },
    {
      "epoch": 0.2194723959749796,
      "grad_norm": 0.396484375,
      "learning_rate": 1.9143675135393937e-05,
      "loss": 0.6707,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1614
    },
    {
      "epoch": 0.2196083763937993,
      "grad_norm": 0.3359375,
      "learning_rate": 1.9141751807099006e-05,
      "loss": 0.6713,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1615
    },
    {
      "epoch": 0.21974435681261897,
      "grad_norm": 0.353515625,
      "learning_rate": 1.913982641814145e-05,
      "loss": 0.5851,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1616
    },
    {
      "epoch": 0.21988033723143868,
      "grad_norm": 0.3515625,
      "learning_rate": 1.9137898968955276e-05,
      "loss": 0.4886,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1617
    },
    {
      "epoch": 0.22001631765025836,
      "grad_norm": 0.57421875,
      "learning_rate": 1.9135969459974953e-05,
      "loss": 0.569,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1618
    },
    {
      "epoch": 0.22015229806907805,
      "grad_norm": 0.34375,
      "learning_rate": 1.9134037891635417e-05,
      "loss": 0.647,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1619
    },
    {
      "epoch": 0.22028827848789775,
      "grad_norm": 0.53515625,
      "learning_rate": 1.9132104264372065e-05,
      "loss": 0.3506,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1620
    },
    {
      "epoch": 0.22042425890671744,
      "grad_norm": 0.37109375,
      "learning_rate": 1.9130168578620764e-05,
      "loss": 0.5728,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1621
    },
    {
      "epoch": 0.22056023932553712,
      "grad_norm": 0.451171875,
      "learning_rate": 1.9128230834817837e-05,
      "loss": 0.6652,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1622
    },
    {
      "epoch": 0.22069621974435683,
      "grad_norm": 0.283203125,
      "learning_rate": 1.912629103340008e-05,
      "loss": 0.6055,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1623
    },
    {
      "epoch": 0.2208322001631765,
      "grad_norm": 0.65625,
      "learning_rate": 1.912434917480474e-05,
      "loss": 0.726,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1624
    },
    {
      "epoch": 0.2209681805819962,
      "grad_norm": 0.294921875,
      "learning_rate": 1.9122405259469548e-05,
      "loss": 0.5804,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1625
    },
    {
      "epoch": 0.22110416100081587,
      "grad_norm": 0.400390625,
      "learning_rate": 1.9120459287832677e-05,
      "loss": 0.5794,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1626
    },
    {
      "epoch": 0.22124014141963558,
      "grad_norm": 0.337890625,
      "learning_rate": 1.911851126033277e-05,
      "loss": 0.5417,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1627
    },
    {
      "epoch": 0.22137612183845526,
      "grad_norm": 0.49609375,
      "learning_rate": 1.911656117740895e-05,
      "loss": 0.777,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1628
    },
    {
      "epoch": 0.22151210225727494,
      "grad_norm": 0.75390625,
      "learning_rate": 1.9114609039500775e-05,
      "loss": 0.8013,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1629
    },
    {
      "epoch": 0.22164808267609465,
      "grad_norm": 0.482421875,
      "learning_rate": 1.9112654847048287e-05,
      "loss": 0.6296,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1630
    },
    {
      "epoch": 0.22178406309491433,
      "grad_norm": 0.400390625,
      "learning_rate": 1.9110698600491985e-05,
      "loss": 0.7471,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1631
    },
    {
      "epoch": 0.221920043513734,
      "grad_norm": 0.57421875,
      "learning_rate": 1.9108740300272832e-05,
      "loss": 0.4813,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1632
    },
    {
      "epoch": 0.22205602393255372,
      "grad_norm": 0.78125,
      "learning_rate": 1.910677994683225e-05,
      "loss": 0.937,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1633
    },
    {
      "epoch": 0.2221920043513734,
      "grad_norm": 0.53125,
      "learning_rate": 1.9104817540612125e-05,
      "loss": 0.7738,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1634
    },
    {
      "epoch": 0.22232798477019308,
      "grad_norm": 0.369140625,
      "learning_rate": 1.9102853082054817e-05,
      "loss": 0.7316,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1635
    },
    {
      "epoch": 0.2224639651890128,
      "grad_norm": 0.318359375,
      "learning_rate": 1.9100886571603126e-05,
      "loss": 0.6683,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1636
    },
    {
      "epoch": 0.22259994560783247,
      "grad_norm": 0.296875,
      "learning_rate": 1.9098918009700337e-05,
      "loss": 0.4448,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1637
    },
    {
      "epoch": 0.22273592602665215,
      "grad_norm": 1.6875,
      "learning_rate": 1.909694739679018e-05,
      "loss": 0.7692,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1638
    },
    {
      "epoch": 0.22287190644547186,
      "grad_norm": 0.24609375,
      "learning_rate": 1.9094974733316865e-05,
      "loss": 0.5068,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1639
    },
    {
      "epoch": 0.22300788686429154,
      "grad_norm": 0.62109375,
      "learning_rate": 1.9093000019725044e-05,
      "loss": 0.6294,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1640
    },
    {
      "epoch": 0.22314386728311122,
      "grad_norm": 0.3203125,
      "learning_rate": 1.9091023256459852e-05,
      "loss": 0.5765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1641
    },
    {
      "epoch": 0.22327984770193093,
      "grad_norm": 0.310546875,
      "learning_rate": 1.9089044443966868e-05,
      "loss": 0.5675,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1642
    },
    {
      "epoch": 0.22341582812075061,
      "grad_norm": 0.314453125,
      "learning_rate": 1.9087063582692143e-05,
      "loss": 0.5291,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1643
    },
    {
      "epoch": 0.2235518085395703,
      "grad_norm": 0.310546875,
      "learning_rate": 1.9085080673082185e-05,
      "loss": 0.6312,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1644
    },
    {
      "epoch": 0.22368778895839,
      "grad_norm": 0.6015625,
      "learning_rate": 1.908309571558397e-05,
      "loss": 0.7936,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1645
    },
    {
      "epoch": 0.22382376937720969,
      "grad_norm": 0.36328125,
      "learning_rate": 1.9081108710644933e-05,
      "loss": 0.7062,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1646
    },
    {
      "epoch": 0.22395974979602937,
      "grad_norm": 3.640625,
      "learning_rate": 1.9079119658712963e-05,
      "loss": 0.9623,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1647
    },
    {
      "epoch": 0.22409573021484905,
      "grad_norm": 0.404296875,
      "learning_rate": 1.907712856023642e-05,
      "loss": 0.6563,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1648
    },
    {
      "epoch": 0.22423171063366876,
      "grad_norm": 0.46484375,
      "learning_rate": 1.9075135415664125e-05,
      "loss": 0.686,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1649
    },
    {
      "epoch": 0.22436769105248844,
      "grad_norm": 0.57421875,
      "learning_rate": 1.9073140225445352e-05,
      "loss": 0.6377,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1650
    },
    {
      "epoch": 0.22450367147130812,
      "grad_norm": 0.515625,
      "learning_rate": 1.907114299002984e-05,
      "loss": 0.8743,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1651
    },
    {
      "epoch": 0.22463965189012783,
      "grad_norm": 0.427734375,
      "learning_rate": 1.9069143709867797e-05,
      "loss": 0.8509,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1652
    },
    {
      "epoch": 0.2247756323089475,
      "grad_norm": 0.458984375,
      "learning_rate": 1.906714238540988e-05,
      "loss": 0.7132,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1653
    },
    {
      "epoch": 0.2249116127277672,
      "grad_norm": 0.875,
      "learning_rate": 1.9065139017107218e-05,
      "loss": 0.8034,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1654
    },
    {
      "epoch": 0.2250475931465869,
      "grad_norm": 0.8203125,
      "learning_rate": 1.906313360541139e-05,
      "loss": 0.63,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1655
    },
    {
      "epoch": 0.22518357356540658,
      "grad_norm": 0.423828125,
      "learning_rate": 1.9061126150774437e-05,
      "loss": 0.7083,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1656
    },
    {
      "epoch": 0.22531955398422626,
      "grad_norm": 0.890625,
      "learning_rate": 1.9059116653648867e-05,
      "loss": 0.8542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1657
    },
    {
      "epoch": 0.22545553440304597,
      "grad_norm": 0.447265625,
      "learning_rate": 1.9057105114487648e-05,
      "loss": 0.7384,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1658
    },
    {
      "epoch": 0.22559151482186565,
      "grad_norm": 0.26171875,
      "learning_rate": 1.9055091533744204e-05,
      "loss": 0.4655,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1659
    },
    {
      "epoch": 0.22572749524068533,
      "grad_norm": 0.57421875,
      "learning_rate": 1.9053075911872415e-05,
      "loss": 0.799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1660
    },
    {
      "epoch": 0.22586347565950504,
      "grad_norm": 0.431640625,
      "learning_rate": 1.9051058249326638e-05,
      "loss": 0.7148,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1661
    },
    {
      "epoch": 0.22599945607832472,
      "grad_norm": 0.56640625,
      "learning_rate": 1.9049038546561667e-05,
      "loss": 0.6167,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1662
    },
    {
      "epoch": 0.2261354364971444,
      "grad_norm": 0.640625,
      "learning_rate": 1.904701680403278e-05,
      "loss": 0.5216,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1663
    },
    {
      "epoch": 0.2262714169159641,
      "grad_norm": 0.5078125,
      "learning_rate": 1.904499302219569e-05,
      "loss": 0.5946,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1664
    },
    {
      "epoch": 0.2264073973347838,
      "grad_norm": 0.392578125,
      "learning_rate": 1.9042967201506594e-05,
      "loss": 0.6587,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1665
    },
    {
      "epoch": 0.22654337775360348,
      "grad_norm": 1.0078125,
      "learning_rate": 1.9040939342422128e-05,
      "loss": 0.6885,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1666
    },
    {
      "epoch": 0.22667935817242318,
      "grad_norm": 0.322265625,
      "learning_rate": 1.9038909445399397e-05,
      "loss": 0.5942,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1667
    },
    {
      "epoch": 0.22681533859124287,
      "grad_norm": 0.421875,
      "learning_rate": 1.9036877510895967e-05,
      "loss": 0.5425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1668
    },
    {
      "epoch": 0.22695131901006255,
      "grad_norm": 0.310546875,
      "learning_rate": 1.903484353936986e-05,
      "loss": 0.6292,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1669
    },
    {
      "epoch": 0.22708729942888223,
      "grad_norm": 0.25390625,
      "learning_rate": 1.903280753127956e-05,
      "loss": 0.411,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1670
    },
    {
      "epoch": 0.22722327984770194,
      "grad_norm": 0.283203125,
      "learning_rate": 1.9030769487084005e-05,
      "loss": 0.5378,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1671
    },
    {
      "epoch": 0.22735926026652162,
      "grad_norm": 0.44921875,
      "learning_rate": 1.9028729407242598e-05,
      "loss": 0.8638,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1672
    },
    {
      "epoch": 0.2274952406853413,
      "grad_norm": 0.59765625,
      "learning_rate": 1.90266872922152e-05,
      "loss": 0.6331,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1673
    },
    {
      "epoch": 0.227631221104161,
      "grad_norm": 0.3671875,
      "learning_rate": 1.902464314246212e-05,
      "loss": 0.6816,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1674
    },
    {
      "epoch": 0.2277672015229807,
      "grad_norm": 0.322265625,
      "learning_rate": 1.902259695844414e-05,
      "loss": 0.5373,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1675
    },
    {
      "epoch": 0.22790318194180037,
      "grad_norm": 0.322265625,
      "learning_rate": 1.90205487406225e-05,
      "loss": 0.5334,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1676
    },
    {
      "epoch": 0.22803916236062008,
      "grad_norm": 0.306640625,
      "learning_rate": 1.9018498489458882e-05,
      "loss": 0.5456,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1677
    },
    {
      "epoch": 0.22817514277943976,
      "grad_norm": 0.412109375,
      "learning_rate": 1.9016446205415446e-05,
      "loss": 0.6628,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1678
    },
    {
      "epoch": 0.22831112319825944,
      "grad_norm": 0.6796875,
      "learning_rate": 1.90143918889548e-05,
      "loss": 0.6764,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1679
    },
    {
      "epoch": 0.22844710361707915,
      "grad_norm": 0.392578125,
      "learning_rate": 1.901233554054001e-05,
      "loss": 0.6965,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1680
    },
    {
      "epoch": 0.22858308403589883,
      "grad_norm": 0.298828125,
      "learning_rate": 1.9010277160634605e-05,
      "loss": 0.5182,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1681
    },
    {
      "epoch": 0.2287190644547185,
      "grad_norm": 0.33203125,
      "learning_rate": 1.900821674970257e-05,
      "loss": 0.6538,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1682
    },
    {
      "epoch": 0.22885504487353822,
      "grad_norm": 0.3828125,
      "learning_rate": 1.900615430820834e-05,
      "loss": 0.6374,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1683
    },
    {
      "epoch": 0.2289910252923579,
      "grad_norm": 0.4765625,
      "learning_rate": 1.900408983661682e-05,
      "loss": 0.6432,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1684
    },
    {
      "epoch": 0.22912700571117758,
      "grad_norm": 0.458984375,
      "learning_rate": 1.9002023335393366e-05,
      "loss": 0.9085,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1685
    },
    {
      "epoch": 0.2292629861299973,
      "grad_norm": 0.5390625,
      "learning_rate": 1.8999954805003794e-05,
      "loss": 0.7752,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1686
    },
    {
      "epoch": 0.22939896654881697,
      "grad_norm": 0.3828125,
      "learning_rate": 1.8997884245914372e-05,
      "loss": 0.6475,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1687
    },
    {
      "epoch": 0.22953494696763665,
      "grad_norm": 0.36328125,
      "learning_rate": 1.8995811658591836e-05,
      "loss": 0.5662,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1688
    },
    {
      "epoch": 0.22967092738645636,
      "grad_norm": 0.4765625,
      "learning_rate": 1.8993737043503367e-05,
      "loss": 0.7453,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1689
    },
    {
      "epoch": 0.22980690780527604,
      "grad_norm": 0.48828125,
      "learning_rate": 1.899166040111661e-05,
      "loss": 0.6637,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1690
    },
    {
      "epoch": 0.22994288822409573,
      "grad_norm": 0.42578125,
      "learning_rate": 1.8989581731899665e-05,
      "loss": 0.6637,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1691
    },
    {
      "epoch": 0.2300788686429154,
      "grad_norm": 0.46875,
      "learning_rate": 1.898750103632109e-05,
      "loss": 0.7559,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1692
    },
    {
      "epoch": 0.23021484906173512,
      "grad_norm": 0.419921875,
      "learning_rate": 1.89854183148499e-05,
      "loss": 0.6558,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1693
    },
    {
      "epoch": 0.2303508294805548,
      "grad_norm": 0.72265625,
      "learning_rate": 1.898333356795557e-05,
      "loss": 0.7474,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1694
    },
    {
      "epoch": 0.23048680989937448,
      "grad_norm": 0.435546875,
      "learning_rate": 1.8981246796108018e-05,
      "loss": 0.659,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1695
    },
    {
      "epoch": 0.2306227903181942,
      "grad_norm": 0.35546875,
      "learning_rate": 1.8979157999777635e-05,
      "loss": 0.6297,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1696
    },
    {
      "epoch": 0.23075877073701387,
      "grad_norm": 0.66015625,
      "learning_rate": 1.8977067179435254e-05,
      "loss": 0.6324,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1697
    },
    {
      "epoch": 0.23089475115583355,
      "grad_norm": 0.333984375,
      "learning_rate": 1.897497433555218e-05,
      "loss": 0.6437,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1698
    },
    {
      "epoch": 0.23103073157465326,
      "grad_norm": 0.369140625,
      "learning_rate": 1.8972879468600163e-05,
      "loss": 0.5862,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1699
    },
    {
      "epoch": 0.23116671199347294,
      "grad_norm": 0.51171875,
      "learning_rate": 1.897078257905141e-05,
      "loss": 0.8094,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1700
    },
    {
      "epoch": 0.23130269241229262,
      "grad_norm": 0.392578125,
      "learning_rate": 1.8968683667378588e-05,
      "loss": 0.7062,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1701
    },
    {
      "epoch": 0.23143867283111233,
      "grad_norm": 0.28515625,
      "learning_rate": 1.896658273405481e-05,
      "loss": 0.5568,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1702
    },
    {
      "epoch": 0.231574653249932,
      "grad_norm": 0.54296875,
      "learning_rate": 1.8964479779553665e-05,
      "loss": 0.6567,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1703
    },
    {
      "epoch": 0.2317106336687517,
      "grad_norm": 0.29296875,
      "learning_rate": 1.896237480434917e-05,
      "loss": 0.3986,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1704
    },
    {
      "epoch": 0.2318466140875714,
      "grad_norm": 0.384765625,
      "learning_rate": 1.8960267808915823e-05,
      "loss": 0.5697,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1705
    },
    {
      "epoch": 0.23198259450639108,
      "grad_norm": 0.4609375,
      "learning_rate": 1.8958158793728563e-05,
      "loss": 0.806,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1706
    },
    {
      "epoch": 0.23211857492521076,
      "grad_norm": 0.318359375,
      "learning_rate": 1.8956047759262788e-05,
      "loss": 0.5067,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1707
    },
    {
      "epoch": 0.23225455534403047,
      "grad_norm": 0.5390625,
      "learning_rate": 1.895393470599435e-05,
      "loss": 0.9036,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1708
    },
    {
      "epoch": 0.23239053576285015,
      "grad_norm": 0.3359375,
      "learning_rate": 1.895181963439956e-05,
      "loss": 0.6538,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1709
    },
    {
      "epoch": 0.23252651618166983,
      "grad_norm": 0.453125,
      "learning_rate": 1.8949702544955175e-05,
      "loss": 0.5877,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1710
    },
    {
      "epoch": 0.23266249660048952,
      "grad_norm": 0.203125,
      "learning_rate": 1.894758343813842e-05,
      "loss": 0.4023,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1711
    },
    {
      "epoch": 0.23279847701930922,
      "grad_norm": 0.6015625,
      "learning_rate": 1.8945462314426958e-05,
      "loss": 0.7729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1712
    },
    {
      "epoch": 0.2329344574381289,
      "grad_norm": 0.34375,
      "learning_rate": 1.894333917429893e-05,
      "loss": 0.6763,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1713
    },
    {
      "epoch": 0.2330704378569486,
      "grad_norm": 0.40625,
      "learning_rate": 1.8941214018232905e-05,
      "loss": 0.8161,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1714
    },
    {
      "epoch": 0.2332064182757683,
      "grad_norm": 0.326171875,
      "learning_rate": 1.893908684670793e-05,
      "loss": 0.4992,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1715
    },
    {
      "epoch": 0.23334239869458798,
      "grad_norm": 0.47265625,
      "learning_rate": 1.8936957660203484e-05,
      "loss": 0.6032,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1716
    },
    {
      "epoch": 0.23347837911340766,
      "grad_norm": 0.90234375,
      "learning_rate": 1.8934826459199517e-05,
      "loss": 0.8337,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1717
    },
    {
      "epoch": 0.23361435953222737,
      "grad_norm": 0.486328125,
      "learning_rate": 1.893269324417643e-05,
      "loss": 0.7456,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1718
    },
    {
      "epoch": 0.23375033995104705,
      "grad_norm": 0.64453125,
      "learning_rate": 1.8930558015615076e-05,
      "loss": 0.4889,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1719
    },
    {
      "epoch": 0.23388632036986673,
      "grad_norm": 0.50390625,
      "learning_rate": 1.8928420773996757e-05,
      "loss": 0.6263,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1720
    },
    {
      "epoch": 0.23402230078868644,
      "grad_norm": 0.40234375,
      "learning_rate": 1.892628151980324e-05,
      "loss": 0.5957,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1721
    },
    {
      "epoch": 0.23415828120750612,
      "grad_norm": 0.33203125,
      "learning_rate": 1.892414025351673e-05,
      "loss": 0.557,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1722
    },
    {
      "epoch": 0.2342942616263258,
      "grad_norm": 0.46875,
      "learning_rate": 1.8921996975619903e-05,
      "loss": 0.8913,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1723
    },
    {
      "epoch": 0.2344302420451455,
      "grad_norm": 0.361328125,
      "learning_rate": 1.8919851686595875e-05,
      "loss": 0.653,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1724
    },
    {
      "epoch": 0.2345662224639652,
      "grad_norm": 0.361328125,
      "learning_rate": 1.8917704386928223e-05,
      "loss": 0.7302,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1725
    },
    {
      "epoch": 0.23470220288278487,
      "grad_norm": 0.28515625,
      "learning_rate": 1.891555507710097e-05,
      "loss": 0.4757,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1726
    },
    {
      "epoch": 0.23483818330160458,
      "grad_norm": 0.388671875,
      "learning_rate": 1.8913403757598603e-05,
      "loss": 0.5924,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1727
    },
    {
      "epoch": 0.23497416372042426,
      "grad_norm": 0.376953125,
      "learning_rate": 1.8911250428906056e-05,
      "loss": 0.6265,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1728
    },
    {
      "epoch": 0.23511014413924394,
      "grad_norm": 0.52734375,
      "learning_rate": 1.8909095091508708e-05,
      "loss": 0.7531,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1729
    },
    {
      "epoch": 0.23524612455806365,
      "grad_norm": 0.52734375,
      "learning_rate": 1.8906937745892407e-05,
      "loss": 0.7705,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1730
    },
    {
      "epoch": 0.23538210497688333,
      "grad_norm": 0.625,
      "learning_rate": 1.8904778392543435e-05,
      "loss": 0.8866,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1731
    },
    {
      "epoch": 0.23551808539570301,
      "grad_norm": 0.8125,
      "learning_rate": 1.8902617031948552e-05,
      "loss": 0.7661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1732
    },
    {
      "epoch": 0.2356540658145227,
      "grad_norm": 0.546875,
      "learning_rate": 1.8900453664594937e-05,
      "loss": 0.5744,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1733
    },
    {
      "epoch": 0.2357900462333424,
      "grad_norm": 0.439453125,
      "learning_rate": 1.8898288290970256e-05,
      "loss": 0.7648,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1734
    },
    {
      "epoch": 0.23592602665216209,
      "grad_norm": 0.3046875,
      "learning_rate": 1.8896120911562595e-05,
      "loss": 0.4618,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1735
    },
    {
      "epoch": 0.23606200707098177,
      "grad_norm": 0.58984375,
      "learning_rate": 1.8893951526860522e-05,
      "loss": 0.6603,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1736
    },
    {
      "epoch": 0.23619798748980148,
      "grad_norm": 0.3828125,
      "learning_rate": 1.8891780137353036e-05,
      "loss": 0.5915,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1737
    },
    {
      "epoch": 0.23633396790862116,
      "grad_norm": 0.81640625,
      "learning_rate": 1.8889606743529593e-05,
      "loss": 0.6719,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1738
    },
    {
      "epoch": 0.23646994832744084,
      "grad_norm": 0.3828125,
      "learning_rate": 1.8887431345880107e-05,
      "loss": 0.5836,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1739
    },
    {
      "epoch": 0.23660592874626055,
      "grad_norm": 0.376953125,
      "learning_rate": 1.8885253944894938e-05,
      "loss": 0.6799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1740
    },
    {
      "epoch": 0.23674190916508023,
      "grad_norm": 0.37109375,
      "learning_rate": 1.8883074541064895e-05,
      "loss": 0.4876,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1741
    },
    {
      "epoch": 0.2368778895838999,
      "grad_norm": 0.3515625,
      "learning_rate": 1.8880893134881247e-05,
      "loss": 0.7945,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1742
    },
    {
      "epoch": 0.23701387000271962,
      "grad_norm": 0.470703125,
      "learning_rate": 1.8878709726835707e-05,
      "loss": 0.7798,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1743
    },
    {
      "epoch": 0.2371498504215393,
      "grad_norm": 0.28515625,
      "learning_rate": 1.887652431742044e-05,
      "loss": 0.6087,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1744
    },
    {
      "epoch": 0.23728583084035898,
      "grad_norm": 0.302734375,
      "learning_rate": 1.8874336907128073e-05,
      "loss": 0.4906,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1745
    },
    {
      "epoch": 0.2374218112591787,
      "grad_norm": 0.421875,
      "learning_rate": 1.8872147496451666e-05,
      "loss": 0.7384,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1746
    },
    {
      "epoch": 0.23755779167799837,
      "grad_norm": 0.330078125,
      "learning_rate": 1.8869956085884743e-05,
      "loss": 0.7183,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1747
    },
    {
      "epoch": 0.23769377209681805,
      "grad_norm": 0.62890625,
      "learning_rate": 1.8867762675921272e-05,
      "loss": 0.6781,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1748
    },
    {
      "epoch": 0.23782975251563776,
      "grad_norm": 0.287109375,
      "learning_rate": 1.8865567267055676e-05,
      "loss": 0.576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1749
    },
    {
      "epoch": 0.23796573293445744,
      "grad_norm": 0.478515625,
      "learning_rate": 1.8863369859782824e-05,
      "loss": 0.9368,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1750
    },
    {
      "epoch": 0.23810171335327712,
      "grad_norm": 0.369140625,
      "learning_rate": 1.8861170454598046e-05,
      "loss": 0.4471,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1751
    },
    {
      "epoch": 0.23823769377209683,
      "grad_norm": 0.8203125,
      "learning_rate": 1.885896905199711e-05,
      "loss": 0.8625,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1752
    },
    {
      "epoch": 0.2383736741909165,
      "grad_norm": 0.337890625,
      "learning_rate": 1.8856765652476235e-05,
      "loss": 0.6043,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1753
    },
    {
      "epoch": 0.2385096546097362,
      "grad_norm": 0.388671875,
      "learning_rate": 1.8854560256532098e-05,
      "loss": 0.6655,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1754
    },
    {
      "epoch": 0.23864563502855587,
      "grad_norm": 0.64453125,
      "learning_rate": 1.885235286466183e-05,
      "loss": 0.7871,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1755
    },
    {
      "epoch": 0.23878161544737558,
      "grad_norm": 0.326171875,
      "learning_rate": 1.885014347736299e-05,
      "loss": 0.6213,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1756
    },
    {
      "epoch": 0.23891759586619526,
      "grad_norm": 0.302734375,
      "learning_rate": 1.884793209513361e-05,
      "loss": 0.4642,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1757
    },
    {
      "epoch": 0.23905357628501495,
      "grad_norm": 0.388671875,
      "learning_rate": 1.884571871847216e-05,
      "loss": 0.6995,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1758
    },
    {
      "epoch": 0.23918955670383465,
      "grad_norm": 0.283203125,
      "learning_rate": 1.8843503347877565e-05,
      "loss": 0.5578,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1759
    },
    {
      "epoch": 0.23932553712265434,
      "grad_norm": 0.5,
      "learning_rate": 1.884128598384919e-05,
      "loss": 0.7852,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1760
    },
    {
      "epoch": 0.23946151754147402,
      "grad_norm": 0.484375,
      "learning_rate": 1.883906662688686e-05,
      "loss": 0.7333,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1761
    },
    {
      "epoch": 0.23959749796029373,
      "grad_norm": 0.255859375,
      "learning_rate": 1.883684527749085e-05,
      "loss": 0.5519,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1762
    },
    {
      "epoch": 0.2397334783791134,
      "grad_norm": 0.71875,
      "learning_rate": 1.883462193616187e-05,
      "loss": 0.8322,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1763
    },
    {
      "epoch": 0.2398694587979331,
      "grad_norm": 0.546875,
      "learning_rate": 1.8832396603401093e-05,
      "loss": 0.6589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1764
    },
    {
      "epoch": 0.2400054392167528,
      "grad_norm": 0.57421875,
      "learning_rate": 1.8830169279710137e-05,
      "loss": 0.6414,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1765
    },
    {
      "epoch": 0.24014141963557248,
      "grad_norm": 0.419921875,
      "learning_rate": 1.8827939965591068e-05,
      "loss": 0.651,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1766
    },
    {
      "epoch": 0.24027740005439216,
      "grad_norm": 0.333984375,
      "learning_rate": 1.8825708661546396e-05,
      "loss": 0.4673,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1767
    },
    {
      "epoch": 0.24041338047321187,
      "grad_norm": 0.48828125,
      "learning_rate": 1.882347536807909e-05,
      "loss": 0.7371,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1768
    },
    {
      "epoch": 0.24054936089203155,
      "grad_norm": 0.41796875,
      "learning_rate": 1.8821240085692563e-05,
      "loss": 0.7054,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1769
    },
    {
      "epoch": 0.24068534131085123,
      "grad_norm": 0.41015625,
      "learning_rate": 1.881900281489067e-05,
      "loss": 0.7259,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1770
    },
    {
      "epoch": 0.24082132172967094,
      "grad_norm": 0.259765625,
      "learning_rate": 1.8816763556177715e-05,
      "loss": 0.5513,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1771
    },
    {
      "epoch": 0.24095730214849062,
      "grad_norm": 0.416015625,
      "learning_rate": 1.8814522310058465e-05,
      "loss": 0.5228,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1772
    },
    {
      "epoch": 0.2410932825673103,
      "grad_norm": 0.5703125,
      "learning_rate": 1.881227907703812e-05,
      "loss": 0.5731,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1773
    },
    {
      "epoch": 0.24122926298613,
      "grad_norm": 0.36328125,
      "learning_rate": 1.8810033857622328e-05,
      "loss": 0.634,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1774
    },
    {
      "epoch": 0.2413652434049497,
      "grad_norm": 0.5546875,
      "learning_rate": 1.8807786652317194e-05,
      "loss": 0.5021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1775
    },
    {
      "epoch": 0.24150122382376937,
      "grad_norm": 0.55078125,
      "learning_rate": 1.8805537461629266e-05,
      "loss": 0.5768,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1776
    },
    {
      "epoch": 0.24163720424258905,
      "grad_norm": 0.7109375,
      "learning_rate": 1.8803286286065535e-05,
      "loss": 0.7269,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1777
    },
    {
      "epoch": 0.24177318466140876,
      "grad_norm": 1.9140625,
      "learning_rate": 1.8801033126133443e-05,
      "loss": 0.6976,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1778
    },
    {
      "epoch": 0.24190916508022844,
      "grad_norm": 0.451171875,
      "learning_rate": 1.8798777982340884e-05,
      "loss": 0.6878,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1779
    },
    {
      "epoch": 0.24204514549904813,
      "grad_norm": 0.322265625,
      "learning_rate": 1.8796520855196197e-05,
      "loss": 0.4855,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1780
    },
    {
      "epoch": 0.24218112591786783,
      "grad_norm": 2.1875,
      "learning_rate": 1.8794261745208156e-05,
      "loss": 0.8027,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1781
    },
    {
      "epoch": 0.24231710633668752,
      "grad_norm": 0.32421875,
      "learning_rate": 1.8792000652886005e-05,
      "loss": 0.5671,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1782
    },
    {
      "epoch": 0.2424530867555072,
      "grad_norm": 0.48046875,
      "learning_rate": 1.878973757873941e-05,
      "loss": 0.6986,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1783
    },
    {
      "epoch": 0.2425890671743269,
      "grad_norm": 0.52734375,
      "learning_rate": 1.87874725232785e-05,
      "loss": 0.5967,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1784
    },
    {
      "epoch": 0.2427250475931466,
      "grad_norm": 0.50390625,
      "learning_rate": 1.8785205487013853e-05,
      "loss": 0.8584,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1785
    },
    {
      "epoch": 0.24286102801196627,
      "grad_norm": 0.42578125,
      "learning_rate": 1.8782936470456476e-05,
      "loss": 0.7188,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1786
    },
    {
      "epoch": 0.24299700843078598,
      "grad_norm": 0.283203125,
      "learning_rate": 1.8780665474117838e-05,
      "loss": 0.5908,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1787
    },
    {
      "epoch": 0.24313298884960566,
      "grad_norm": 0.64453125,
      "learning_rate": 1.8778392498509848e-05,
      "loss": 0.755,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1788
    },
    {
      "epoch": 0.24326896926842534,
      "grad_norm": 1.1953125,
      "learning_rate": 1.8776117544144866e-05,
      "loss": 0.6968,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1789
    },
    {
      "epoch": 0.24340494968724505,
      "grad_norm": 1.0078125,
      "learning_rate": 1.8773840611535693e-05,
      "loss": 0.6297,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1790
    },
    {
      "epoch": 0.24354093010606473,
      "grad_norm": 0.5625,
      "learning_rate": 1.8771561701195573e-05,
      "loss": 0.7892,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1791
    },
    {
      "epoch": 0.2436769105248844,
      "grad_norm": 0.3828125,
      "learning_rate": 1.8769280813638208e-05,
      "loss": 0.6471,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1792
    },
    {
      "epoch": 0.24381289094370412,
      "grad_norm": 0.337890625,
      "learning_rate": 1.8766997949377726e-05,
      "loss": 0.5863,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1793
    },
    {
      "epoch": 0.2439488713625238,
      "grad_norm": 0.423828125,
      "learning_rate": 1.8764713108928727e-05,
      "loss": 0.8213,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1794
    },
    {
      "epoch": 0.24408485178134348,
      "grad_norm": 0.337890625,
      "learning_rate": 1.876242629280623e-05,
      "loss": 0.5778,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1795
    },
    {
      "epoch": 0.24422083220016316,
      "grad_norm": 0.31640625,
      "learning_rate": 1.8760137501525723e-05,
      "loss": 0.5397,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1796
    },
    {
      "epoch": 0.24435681261898287,
      "grad_norm": 0.640625,
      "learning_rate": 1.8757846735603118e-05,
      "loss": 0.6081,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1797
    },
    {
      "epoch": 0.24449279303780255,
      "grad_norm": 0.427734375,
      "learning_rate": 1.8755553995554786e-05,
      "loss": 0.7645,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1798
    },
    {
      "epoch": 0.24462877345662223,
      "grad_norm": 0.43359375,
      "learning_rate": 1.8753259281897532e-05,
      "loss": 0.6867,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1799
    },
    {
      "epoch": 0.24476475387544194,
      "grad_norm": 0.5859375,
      "learning_rate": 1.8750962595148622e-05,
      "loss": 0.9146,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1800
    },
    {
      "epoch": 0.24490073429426162,
      "grad_norm": 0.373046875,
      "learning_rate": 1.8748663935825755e-05,
      "loss": 0.6917,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1801
    },
    {
      "epoch": 0.2450367147130813,
      "grad_norm": 0.408203125,
      "learning_rate": 1.8746363304447073e-05,
      "loss": 0.7674,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1802
    },
    {
      "epoch": 0.24517269513190101,
      "grad_norm": 0.390625,
      "learning_rate": 1.874406070153117e-05,
      "loss": 0.5991,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1803
    },
    {
      "epoch": 0.2453086755507207,
      "grad_norm": 0.318359375,
      "learning_rate": 1.8741756127597074e-05,
      "loss": 0.6011,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1804
    },
    {
      "epoch": 0.24544465596954038,
      "grad_norm": 0.3671875,
      "learning_rate": 1.8739449583164274e-05,
      "loss": 0.5887,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1805
    },
    {
      "epoch": 0.24558063638836009,
      "grad_norm": 0.353515625,
      "learning_rate": 1.8737141068752688e-05,
      "loss": 0.5607,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1806
    },
    {
      "epoch": 0.24571661680717977,
      "grad_norm": 0.484375,
      "learning_rate": 1.8734830584882685e-05,
      "loss": 0.7088,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1807
    },
    {
      "epoch": 0.24585259722599945,
      "grad_norm": 0.59375,
      "learning_rate": 1.8732518132075072e-05,
      "loss": 0.7363,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1808
    },
    {
      "epoch": 0.24598857764481916,
      "grad_norm": 0.32421875,
      "learning_rate": 1.873020371085111e-05,
      "loss": 0.6258,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1809
    },
    {
      "epoch": 0.24612455806363884,
      "grad_norm": 0.458984375,
      "learning_rate": 1.8727887321732497e-05,
      "loss": 0.9084,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1810
    },
    {
      "epoch": 0.24626053848245852,
      "grad_norm": 0.5234375,
      "learning_rate": 1.872556896524137e-05,
      "loss": 0.6825,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1811
    },
    {
      "epoch": 0.24639651890127823,
      "grad_norm": 0.51953125,
      "learning_rate": 1.872324864190032e-05,
      "loss": 0.401,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1812
    },
    {
      "epoch": 0.2465324993200979,
      "grad_norm": 0.41796875,
      "learning_rate": 1.8720926352232376e-05,
      "loss": 0.5555,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1813
    },
    {
      "epoch": 0.2466684797389176,
      "grad_norm": 0.4296875,
      "learning_rate": 1.8718602096761008e-05,
      "loss": 0.6025,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1814
    },
    {
      "epoch": 0.2468044601577373,
      "grad_norm": 0.451171875,
      "learning_rate": 1.8716275876010135e-05,
      "loss": 0.6206,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1815
    },
    {
      "epoch": 0.24694044057655698,
      "grad_norm": 0.33203125,
      "learning_rate": 1.8713947690504114e-05,
      "loss": 0.6121,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1816
    },
    {
      "epoch": 0.24707642099537666,
      "grad_norm": 0.455078125,
      "learning_rate": 1.8711617540767743e-05,
      "loss": 0.6143,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1817
    },
    {
      "epoch": 0.24721240141419634,
      "grad_norm": 0.66796875,
      "learning_rate": 1.870928542732627e-05,
      "loss": 0.5825,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1818
    },
    {
      "epoch": 0.24734838183301605,
      "grad_norm": 0.5859375,
      "learning_rate": 1.8706951350705384e-05,
      "loss": 0.7082,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1819
    },
    {
      "epoch": 0.24748436225183573,
      "grad_norm": 0.6484375,
      "learning_rate": 1.8704615311431214e-05,
      "loss": 0.6488,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1820
    },
    {
      "epoch": 0.2476203426706554,
      "grad_norm": 0.3046875,
      "learning_rate": 1.8702277310030324e-05,
      "loss": 0.5785,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1821
    },
    {
      "epoch": 0.24775632308947512,
      "grad_norm": 0.51953125,
      "learning_rate": 1.8699937347029735e-05,
      "loss": 0.4901,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1822
    },
    {
      "epoch": 0.2478923035082948,
      "grad_norm": 0.376953125,
      "learning_rate": 1.86975954229569e-05,
      "loss": 0.6455,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1823
    },
    {
      "epoch": 0.24802828392711448,
      "grad_norm": 0.47265625,
      "learning_rate": 1.8695251538339724e-05,
      "loss": 0.7453,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1824
    },
    {
      "epoch": 0.2481642643459342,
      "grad_norm": 0.263671875,
      "learning_rate": 1.869290569370654e-05,
      "loss": 0.4228,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1825
    },
    {
      "epoch": 0.24830024476475387,
      "grad_norm": 0.34375,
      "learning_rate": 1.869055788958613e-05,
      "loss": 0.5335,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1826
    },
    {
      "epoch": 0.24843622518357356,
      "grad_norm": 0.51953125,
      "learning_rate": 1.8688208126507725e-05,
      "loss": 0.7948,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1827
    },
    {
      "epoch": 0.24857220560239326,
      "grad_norm": 0.76953125,
      "learning_rate": 1.8685856405000984e-05,
      "loss": 0.6579,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1828
    },
    {
      "epoch": 0.24870818602121295,
      "grad_norm": 0.294921875,
      "learning_rate": 1.8683502725596013e-05,
      "loss": 0.5711,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1829
    },
    {
      "epoch": 0.24884416644003263,
      "grad_norm": 0.5546875,
      "learning_rate": 1.8681147088823363e-05,
      "loss": 0.7184,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1830
    },
    {
      "epoch": 0.24898014685885234,
      "grad_norm": 0.396484375,
      "learning_rate": 1.8678789495214026e-05,
      "loss": 0.8141,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1831
    },
    {
      "epoch": 0.24911612727767202,
      "grad_norm": 0.3203125,
      "learning_rate": 1.8676429945299427e-05,
      "loss": 0.6325,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1832
    },
    {
      "epoch": 0.2492521076964917,
      "grad_norm": 0.63671875,
      "learning_rate": 1.867406843961144e-05,
      "loss": 0.7562,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1833
    },
    {
      "epoch": 0.2493880881153114,
      "grad_norm": 0.52734375,
      "learning_rate": 1.8671704978682376e-05,
      "loss": 0.6418,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1834
    },
    {
      "epoch": 0.2495240685341311,
      "grad_norm": 0.384765625,
      "learning_rate": 1.8669339563044987e-05,
      "loss": 0.5983,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1835
    },
    {
      "epoch": 0.24966004895295077,
      "grad_norm": 0.8125,
      "learning_rate": 1.866697219323247e-05,
      "loss": 0.939,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1836
    },
    {
      "epoch": 0.24979602937177048,
      "grad_norm": 0.271484375,
      "learning_rate": 1.866460286977846e-05,
      "loss": 0.41,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1837
    },
    {
      "epoch": 0.24993200979059016,
      "grad_norm": 0.263671875,
      "learning_rate": 1.8662231593217028e-05,
      "loss": 0.4564,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1838
    },
    {
      "epoch": 0.25006799020940984,
      "grad_norm": 0.4296875,
      "learning_rate": 1.8659858364082688e-05,
      "loss": 0.561,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1839
    },
    {
      "epoch": 0.2502039706282295,
      "grad_norm": 1.0625,
      "learning_rate": 1.8657483182910398e-05,
      "loss": 0.7266,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1840
    },
    {
      "epoch": 0.2503399510470492,
      "grad_norm": 0.37890625,
      "learning_rate": 1.865510605023555e-05,
      "loss": 0.717,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1841
    },
    {
      "epoch": 0.25047593146586894,
      "grad_norm": 0.361328125,
      "learning_rate": 1.8652726966593982e-05,
      "loss": 0.6733,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1842
    },
    {
      "epoch": 0.2506119118846886,
      "grad_norm": 0.345703125,
      "learning_rate": 1.8650345932521966e-05,
      "loss": 0.6401,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1843
    },
    {
      "epoch": 0.2507478923035083,
      "grad_norm": 0.47265625,
      "learning_rate": 1.8647962948556217e-05,
      "loss": 0.6836,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1844
    },
    {
      "epoch": 0.250883872722328,
      "grad_norm": 0.453125,
      "learning_rate": 1.864557801523389e-05,
      "loss": 0.7336,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1845
    },
    {
      "epoch": 0.25101985314114766,
      "grad_norm": 0.46875,
      "learning_rate": 1.8643191133092576e-05,
      "loss": 0.6235,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1846
    },
    {
      "epoch": 0.25115583355996735,
      "grad_norm": 0.73046875,
      "learning_rate": 1.8640802302670305e-05,
      "loss": 0.613,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1847
    },
    {
      "epoch": 0.2512918139787871,
      "grad_norm": 0.388671875,
      "learning_rate": 1.8638411524505557e-05,
      "loss": 0.7186,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1848
    },
    {
      "epoch": 0.25142779439760676,
      "grad_norm": 0.90234375,
      "learning_rate": 1.8636018799137234e-05,
      "loss": 0.7969,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1849
    },
    {
      "epoch": 0.25156377481642644,
      "grad_norm": 0.26171875,
      "learning_rate": 1.863362412710469e-05,
      "loss": 0.4289,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1850
    },
    {
      "epoch": 0.2516997552352461,
      "grad_norm": 0.392578125,
      "learning_rate": 1.8631227508947715e-05,
      "loss": 0.6152,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1851
    },
    {
      "epoch": 0.2518357356540658,
      "grad_norm": 0.412109375,
      "learning_rate": 1.8628828945206533e-05,
      "loss": 0.6766,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1852
    },
    {
      "epoch": 0.2519717160728855,
      "grad_norm": 0.34375,
      "learning_rate": 1.8626428436421805e-05,
      "loss": 0.5521,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1853
    },
    {
      "epoch": 0.25210769649170517,
      "grad_norm": 0.66796875,
      "learning_rate": 1.8624025983134643e-05,
      "loss": 0.6899,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1854
    },
    {
      "epoch": 0.2522436769105249,
      "grad_norm": 0.6015625,
      "learning_rate": 1.8621621585886586e-05,
      "loss": 0.8853,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1855
    },
    {
      "epoch": 0.2523796573293446,
      "grad_norm": 0.31640625,
      "learning_rate": 1.8619215245219617e-05,
      "loss": 0.5164,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1856
    },
    {
      "epoch": 0.25251563774816427,
      "grad_norm": 0.45703125,
      "learning_rate": 1.861680696167615e-05,
      "loss": 0.7013,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1857
    },
    {
      "epoch": 0.25265161816698395,
      "grad_norm": 0.291015625,
      "learning_rate": 1.8614396735799044e-05,
      "loss": 0.4188,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1858
    },
    {
      "epoch": 0.25278759858580363,
      "grad_norm": 0.984375,
      "learning_rate": 1.8611984568131595e-05,
      "loss": 0.6956,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1859
    },
    {
      "epoch": 0.2529235790046233,
      "grad_norm": 0.482421875,
      "learning_rate": 1.8609570459217533e-05,
      "loss": 0.5814,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1860
    },
    {
      "epoch": 0.25305955942344305,
      "grad_norm": 0.5625,
      "learning_rate": 1.8607154409601028e-05,
      "loss": 0.8088,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1861
    },
    {
      "epoch": 0.25319553984226273,
      "grad_norm": 0.3125,
      "learning_rate": 1.8604736419826686e-05,
      "loss": 0.4517,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1862
    },
    {
      "epoch": 0.2533315202610824,
      "grad_norm": 0.4375,
      "learning_rate": 1.8602316490439558e-05,
      "loss": 0.6543,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1863
    },
    {
      "epoch": 0.2534675006799021,
      "grad_norm": 0.4140625,
      "learning_rate": 1.8599894621985114e-05,
      "loss": 0.8101,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1864
    },
    {
      "epoch": 0.2536034810987218,
      "grad_norm": 0.45703125,
      "learning_rate": 1.8597470815009285e-05,
      "loss": 0.6048,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1865
    },
    {
      "epoch": 0.25373946151754145,
      "grad_norm": 0.28515625,
      "learning_rate": 1.8595045070058417e-05,
      "loss": 0.5247,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1866
    },
    {
      "epoch": 0.2538754419363612,
      "grad_norm": 0.3515625,
      "learning_rate": 1.8592617387679304e-05,
      "loss": 0.7062,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1867
    },
    {
      "epoch": 0.25401142235518087,
      "grad_norm": 0.345703125,
      "learning_rate": 1.8590187768419184e-05,
      "loss": 0.612,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1868
    },
    {
      "epoch": 0.25414740277400055,
      "grad_norm": 0.7265625,
      "learning_rate": 1.8587756212825713e-05,
      "loss": 0.7669,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1869
    },
    {
      "epoch": 0.25428338319282023,
      "grad_norm": 0.380859375,
      "learning_rate": 1.8585322721446997e-05,
      "loss": 0.5815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1870
    },
    {
      "epoch": 0.2544193636116399,
      "grad_norm": 0.24609375,
      "learning_rate": 1.8582887294831578e-05,
      "loss": 0.41,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1871
    },
    {
      "epoch": 0.2545553440304596,
      "grad_norm": 0.73046875,
      "learning_rate": 1.8580449933528425e-05,
      "loss": 0.7793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1872
    },
    {
      "epoch": 0.2546913244492793,
      "grad_norm": 0.53515625,
      "learning_rate": 1.8578010638086953e-05,
      "loss": 0.742,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1873
    },
    {
      "epoch": 0.254827304868099,
      "grad_norm": 0.34375,
      "learning_rate": 1.8575569409057005e-05,
      "loss": 0.627,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1874
    },
    {
      "epoch": 0.2549632852869187,
      "grad_norm": 0.453125,
      "learning_rate": 1.857312624698887e-05,
      "loss": 0.6976,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1875
    },
    {
      "epoch": 0.2550992657057384,
      "grad_norm": 0.3515625,
      "learning_rate": 1.8570681152433262e-05,
      "loss": 0.5562,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1876
    },
    {
      "epoch": 0.25523524612455806,
      "grad_norm": 0.384765625,
      "learning_rate": 1.8568234125941334e-05,
      "loss": 0.5863,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1877
    },
    {
      "epoch": 0.25537122654337774,
      "grad_norm": 0.8125,
      "learning_rate": 1.856578516806468e-05,
      "loss": 0.728,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1878
    },
    {
      "epoch": 0.2555072069621974,
      "grad_norm": 0.40234375,
      "learning_rate": 1.8563334279355324e-05,
      "loss": 0.7811,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1879
    },
    {
      "epoch": 0.25564318738101716,
      "grad_norm": 0.455078125,
      "learning_rate": 1.8560881460365726e-05,
      "loss": 0.6523,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1880
    },
    {
      "epoch": 0.25577916779983684,
      "grad_norm": 0.337890625,
      "learning_rate": 1.8558426711648777e-05,
      "loss": 0.5686,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1881
    },
    {
      "epoch": 0.2559151482186565,
      "grad_norm": 0.45703125,
      "learning_rate": 1.8555970033757815e-05,
      "loss": 0.7188,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1882
    },
    {
      "epoch": 0.2560511286374762,
      "grad_norm": 0.439453125,
      "learning_rate": 1.85535114272466e-05,
      "loss": 0.7406,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1883
    },
    {
      "epoch": 0.2561871090562959,
      "grad_norm": 0.326171875,
      "learning_rate": 1.8551050892669333e-05,
      "loss": 0.604,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1884
    },
    {
      "epoch": 0.25632308947511556,
      "grad_norm": 0.6328125,
      "learning_rate": 1.8548588430580648e-05,
      "loss": 0.5145,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1885
    },
    {
      "epoch": 0.2564590698939353,
      "grad_norm": 0.310546875,
      "learning_rate": 1.8546124041535617e-05,
      "loss": 0.4775,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1886
    },
    {
      "epoch": 0.256595050312755,
      "grad_norm": 0.3671875,
      "learning_rate": 1.854365772608974e-05,
      "loss": 0.6019,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1887
    },
    {
      "epoch": 0.25673103073157466,
      "grad_norm": 0.39453125,
      "learning_rate": 1.854118948479896e-05,
      "loss": 0.6945,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1888
    },
    {
      "epoch": 0.25686701115039434,
      "grad_norm": 0.35546875,
      "learning_rate": 1.853871931821964e-05,
      "loss": 0.6738,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1889
    },
    {
      "epoch": 0.257002991569214,
      "grad_norm": 0.376953125,
      "learning_rate": 1.8536247226908592e-05,
      "loss": 0.7785,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1890
    },
    {
      "epoch": 0.2571389719880337,
      "grad_norm": 0.359375,
      "learning_rate": 1.8533773211423054e-05,
      "loss": 0.6771,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1891
    },
    {
      "epoch": 0.2572749524068534,
      "grad_norm": 0.384765625,
      "learning_rate": 1.8531297272320705e-05,
      "loss": 0.4513,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1892
    },
    {
      "epoch": 0.2574109328256731,
      "grad_norm": 0.419921875,
      "learning_rate": 1.8528819410159638e-05,
      "loss": 0.6795,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1893
    },
    {
      "epoch": 0.2575469132444928,
      "grad_norm": 0.32421875,
      "learning_rate": 1.8526339625498408e-05,
      "loss": 0.596,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1894
    },
    {
      "epoch": 0.2576828936633125,
      "grad_norm": 0.423828125,
      "learning_rate": 1.852385791889598e-05,
      "loss": 0.7596,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1895
    },
    {
      "epoch": 0.25781887408213217,
      "grad_norm": 0.33203125,
      "learning_rate": 1.8521374290911762e-05,
      "loss": 0.5767,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1896
    },
    {
      "epoch": 0.25795485450095185,
      "grad_norm": 0.31640625,
      "learning_rate": 1.85188887421056e-05,
      "loss": 0.5285,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1897
    },
    {
      "epoch": 0.25809083491977153,
      "grad_norm": 0.412109375,
      "learning_rate": 1.8516401273037757e-05,
      "loss": 0.7995,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1898
    },
    {
      "epoch": 0.25822681533859126,
      "grad_norm": 0.404296875,
      "learning_rate": 1.8513911884268952e-05,
      "loss": 0.6769,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1899
    },
    {
      "epoch": 0.25836279575741095,
      "grad_norm": 0.69921875,
      "learning_rate": 1.851142057636031e-05,
      "loss": 0.689,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1900
    },
    {
      "epoch": 0.2584987761762306,
      "grad_norm": 0.3984375,
      "learning_rate": 1.8508927349873412e-05,
      "loss": 0.7212,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1901
    },
    {
      "epoch": 0.2586347565950503,
      "grad_norm": 0.431640625,
      "learning_rate": 1.850643220537026e-05,
      "loss": 0.5841,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1902
    },
    {
      "epoch": 0.25877073701387,
      "grad_norm": 0.5078125,
      "learning_rate": 1.8503935143413286e-05,
      "loss": 0.6763,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1903
    },
    {
      "epoch": 0.25890671743268967,
      "grad_norm": 0.4296875,
      "learning_rate": 1.8501436164565363e-05,
      "loss": 0.5988,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1904
    },
    {
      "epoch": 0.2590426978515094,
      "grad_norm": 0.5546875,
      "learning_rate": 1.849893526938979e-05,
      "loss": 0.6589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1905
    },
    {
      "epoch": 0.2591786782703291,
      "grad_norm": 0.6015625,
      "learning_rate": 1.8496432458450297e-05,
      "loss": 0.6389,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1906
    },
    {
      "epoch": 0.25931465868914877,
      "grad_norm": 0.333984375,
      "learning_rate": 1.849392773231105e-05,
      "loss": 0.5003,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1907
    },
    {
      "epoch": 0.25945063910796845,
      "grad_norm": 0.400390625,
      "learning_rate": 1.8491421091536643e-05,
      "loss": 0.6304,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1908
    },
    {
      "epoch": 0.25958661952678813,
      "grad_norm": 0.5390625,
      "learning_rate": 1.8488912536692108e-05,
      "loss": 0.5954,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1909
    },
    {
      "epoch": 0.2597225999456078,
      "grad_norm": 0.6953125,
      "learning_rate": 1.84864020683429e-05,
      "loss": 0.6245,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1910
    },
    {
      "epoch": 0.25985858036442755,
      "grad_norm": 0.359375,
      "learning_rate": 1.848388968705491e-05,
      "loss": 0.5407,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1911
    },
    {
      "epoch": 0.25999456078324723,
      "grad_norm": 0.380859375,
      "learning_rate": 1.8481375393394458e-05,
      "loss": 0.4707,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1912
    },
    {
      "epoch": 0.2601305412020669,
      "grad_norm": 0.365234375,
      "learning_rate": 1.84788591879283e-05,
      "loss": 0.7189,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1913
    },
    {
      "epoch": 0.2602665216208866,
      "grad_norm": 0.48046875,
      "learning_rate": 1.847634107122362e-05,
      "loss": 0.8053,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1914
    },
    {
      "epoch": 0.2604025020397063,
      "grad_norm": 0.4375,
      "learning_rate": 1.847382104384803e-05,
      "loss": 0.5955,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1915
    },
    {
      "epoch": 0.26053848245852596,
      "grad_norm": 0.34375,
      "learning_rate": 1.847129910636957e-05,
      "loss": 0.6286,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1916
    },
    {
      "epoch": 0.26067446287734564,
      "grad_norm": 0.3984375,
      "learning_rate": 1.8468775259356725e-05,
      "loss": 0.8094,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1917
    },
    {
      "epoch": 0.2608104432961654,
      "grad_norm": 0.279296875,
      "learning_rate": 1.8466249503378396e-05,
      "loss": 0.491,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1918
    },
    {
      "epoch": 0.26094642371498505,
      "grad_norm": 0.353515625,
      "learning_rate": 1.8463721839003917e-05,
      "loss": 0.6986,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1919
    },
    {
      "epoch": 0.26108240413380474,
      "grad_norm": 0.31640625,
      "learning_rate": 1.846119226680306e-05,
      "loss": 0.4427,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1920
    },
    {
      "epoch": 0.2612183845526244,
      "grad_norm": 0.36328125,
      "learning_rate": 1.845866078734602e-05,
      "loss": 0.8133,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1921
    },
    {
      "epoch": 0.2613543649714441,
      "grad_norm": 0.46875,
      "learning_rate": 1.8456127401203422e-05,
      "loss": 0.5689,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1922
    },
    {
      "epoch": 0.2614903453902638,
      "grad_norm": 0.498046875,
      "learning_rate": 1.845359210894632e-05,
      "loss": 0.7304,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1923
    },
    {
      "epoch": 0.2616263258090835,
      "grad_norm": 0.5390625,
      "learning_rate": 1.8451054911146204e-05,
      "loss": 0.826,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1924
    },
    {
      "epoch": 0.2617623062279032,
      "grad_norm": 0.34765625,
      "learning_rate": 1.8448515808374986e-05,
      "loss": 0.6174,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1925
    },
    {
      "epoch": 0.2618982866467229,
      "grad_norm": 0.54296875,
      "learning_rate": 1.8445974801205017e-05,
      "loss": 0.8062,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1926
    },
    {
      "epoch": 0.26203426706554256,
      "grad_norm": 0.416015625,
      "learning_rate": 1.844343189020906e-05,
      "loss": 0.751,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1927
    },
    {
      "epoch": 0.26217024748436224,
      "grad_norm": 0.515625,
      "learning_rate": 1.844088707596033e-05,
      "loss": 0.8573,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1928
    },
    {
      "epoch": 0.2623062279031819,
      "grad_norm": 0.3671875,
      "learning_rate": 1.843834035903245e-05,
      "loss": 0.6711,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1929
    },
    {
      "epoch": 0.26244220832200166,
      "grad_norm": 0.40625,
      "learning_rate": 1.843579173999949e-05,
      "loss": 0.6681,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1930
    },
    {
      "epoch": 0.26257818874082134,
      "grad_norm": 0.33984375,
      "learning_rate": 1.8433241219435937e-05,
      "loss": 0.6084,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1931
    },
    {
      "epoch": 0.262714169159641,
      "grad_norm": 0.5234375,
      "learning_rate": 1.8430688797916702e-05,
      "loss": 0.7646,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1932
    },
    {
      "epoch": 0.2628501495784607,
      "grad_norm": 0.384765625,
      "learning_rate": 1.8428134476017138e-05,
      "loss": 0.6274,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1933
    },
    {
      "epoch": 0.2629861299972804,
      "grad_norm": 0.3671875,
      "learning_rate": 1.842557825431302e-05,
      "loss": 0.6092,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1934
    },
    {
      "epoch": 0.26312211041610006,
      "grad_norm": 0.32421875,
      "learning_rate": 1.8423020133380555e-05,
      "loss": 0.6562,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1935
    },
    {
      "epoch": 0.26325809083491974,
      "grad_norm": 0.3125,
      "learning_rate": 1.8420460113796367e-05,
      "loss": 0.5887,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1936
    },
    {
      "epoch": 0.2633940712537395,
      "grad_norm": 1.0,
      "learning_rate": 1.8417898196137526e-05,
      "loss": 0.5511,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1937
    },
    {
      "epoch": 0.26353005167255916,
      "grad_norm": 0.7734375,
      "learning_rate": 1.8415334380981507e-05,
      "loss": 0.7542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1938
    },
    {
      "epoch": 0.26366603209137884,
      "grad_norm": 0.4296875,
      "learning_rate": 1.841276866890624e-05,
      "loss": 0.7256,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1939
    },
    {
      "epoch": 0.2638020125101985,
      "grad_norm": 0.318359375,
      "learning_rate": 1.8410201060490053e-05,
      "loss": 0.5348,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1940
    },
    {
      "epoch": 0.2639379929290182,
      "grad_norm": 0.64453125,
      "learning_rate": 1.8407631556311725e-05,
      "loss": 0.5254,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1941
    },
    {
      "epoch": 0.2640739733478379,
      "grad_norm": 0.27734375,
      "learning_rate": 1.8405060156950453e-05,
      "loss": 0.5404,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1942
    },
    {
      "epoch": 0.2642099537666576,
      "grad_norm": 0.322265625,
      "learning_rate": 1.840248686298586e-05,
      "loss": 0.675,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1943
    },
    {
      "epoch": 0.2643459341854773,
      "grad_norm": 0.2890625,
      "learning_rate": 1.8399911674997998e-05,
      "loss": 0.4515,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1944
    },
    {
      "epoch": 0.264481914604297,
      "grad_norm": 0.373046875,
      "learning_rate": 1.8397334593567347e-05,
      "loss": 0.7109,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1945
    },
    {
      "epoch": 0.26461789502311667,
      "grad_norm": 0.6875,
      "learning_rate": 1.8394755619274813e-05,
      "loss": 0.9688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1946
    },
    {
      "epoch": 0.26475387544193635,
      "grad_norm": 0.609375,
      "learning_rate": 1.839217475270173e-05,
      "loss": 0.5249,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1947
    },
    {
      "epoch": 0.26488985586075603,
      "grad_norm": 0.4453125,
      "learning_rate": 1.8389591994429846e-05,
      "loss": 0.6341,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1948
    },
    {
      "epoch": 0.26502583627957577,
      "grad_norm": 0.359375,
      "learning_rate": 1.8387007345041362e-05,
      "loss": 0.6164,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1949
    },
    {
      "epoch": 0.26516181669839545,
      "grad_norm": 0.6640625,
      "learning_rate": 1.8384420805118878e-05,
      "loss": 0.7524,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1950
    },
    {
      "epoch": 0.26529779711721513,
      "grad_norm": 0.5,
      "learning_rate": 1.838183237524544e-05,
      "loss": 0.6087,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1951
    },
    {
      "epoch": 0.2654337775360348,
      "grad_norm": 0.41796875,
      "learning_rate": 1.8379242056004504e-05,
      "loss": 0.71,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1952
    },
    {
      "epoch": 0.2655697579548545,
      "grad_norm": 0.244140625,
      "learning_rate": 1.8376649847979964e-05,
      "loss": 0.4486,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1953
    },
    {
      "epoch": 0.26570573837367417,
      "grad_norm": 0.69921875,
      "learning_rate": 1.8374055751756138e-05,
      "loss": 0.701,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1954
    },
    {
      "epoch": 0.2658417187924939,
      "grad_norm": 0.4453125,
      "learning_rate": 1.8371459767917762e-05,
      "loss": 0.6719,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1955
    },
    {
      "epoch": 0.2659776992113136,
      "grad_norm": 0.80078125,
      "learning_rate": 1.8368861897050002e-05,
      "loss": 0.7104,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1956
    },
    {
      "epoch": 0.26611367963013327,
      "grad_norm": 1.140625,
      "learning_rate": 1.8366262139738454e-05,
      "loss": 0.7358,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1957
    },
    {
      "epoch": 0.26624966004895295,
      "grad_norm": 0.5546875,
      "learning_rate": 1.836366049656913e-05,
      "loss": 0.4793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1958
    },
    {
      "epoch": 0.26638564046777263,
      "grad_norm": 0.37109375,
      "learning_rate": 1.836105696812848e-05,
      "loss": 0.5633,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1959
    },
    {
      "epoch": 0.2665216208865923,
      "grad_norm": 0.37890625,
      "learning_rate": 1.8358451555003363e-05,
      "loss": 0.6291,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1960
    },
    {
      "epoch": 0.266657601305412,
      "grad_norm": 0.359375,
      "learning_rate": 1.8355844257781076e-05,
      "loss": 0.6247,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1961
    },
    {
      "epoch": 0.26679358172423173,
      "grad_norm": 0.4296875,
      "learning_rate": 1.835323507704933e-05,
      "loss": 0.6948,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1962
    },
    {
      "epoch": 0.2669295621430514,
      "grad_norm": 0.60546875,
      "learning_rate": 1.8350624013396276e-05,
      "loss": 0.6621,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1963
    },
    {
      "epoch": 0.2670655425618711,
      "grad_norm": 0.431640625,
      "learning_rate": 1.8348011067410475e-05,
      "loss": 0.8191,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1964
    },
    {
      "epoch": 0.2672015229806908,
      "grad_norm": 0.47265625,
      "learning_rate": 1.834539623968091e-05,
      "loss": 0.6993,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1965
    },
    {
      "epoch": 0.26733750339951046,
      "grad_norm": 0.546875,
      "learning_rate": 1.8342779530797004e-05,
      "loss": 0.8358,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1966
    },
    {
      "epoch": 0.26747348381833014,
      "grad_norm": 0.39453125,
      "learning_rate": 1.8340160941348594e-05,
      "loss": 0.6735,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1967
    },
    {
      "epoch": 0.2676094642371499,
      "grad_norm": 0.451171875,
      "learning_rate": 1.833754047192594e-05,
      "loss": 0.7751,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1968
    },
    {
      "epoch": 0.26774544465596956,
      "grad_norm": 0.33984375,
      "learning_rate": 1.8334918123119724e-05,
      "loss": 0.714,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1969
    },
    {
      "epoch": 0.26788142507478924,
      "grad_norm": 0.4296875,
      "learning_rate": 1.8332293895521062e-05,
      "loss": 0.8368,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1970
    },
    {
      "epoch": 0.2680174054936089,
      "grad_norm": 0.29296875,
      "learning_rate": 1.8329667789721487e-05,
      "loss": 0.5493,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1971
    },
    {
      "epoch": 0.2681533859124286,
      "grad_norm": 0.29296875,
      "learning_rate": 1.832703980631295e-05,
      "loss": 0.458,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1972
    },
    {
      "epoch": 0.2682893663312483,
      "grad_norm": 0.267578125,
      "learning_rate": 1.8324409945887836e-05,
      "loss": 0.488,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1973
    },
    {
      "epoch": 0.268425346750068,
      "grad_norm": 0.62890625,
      "learning_rate": 1.8321778209038944e-05,
      "loss": 0.5677,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1974
    },
    {
      "epoch": 0.2685613271688877,
      "grad_norm": 0.447265625,
      "learning_rate": 1.83191445963595e-05,
      "loss": 0.6532,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1975
    },
    {
      "epoch": 0.2686973075877074,
      "grad_norm": 0.458984375,
      "learning_rate": 1.8316509108443156e-05,
      "loss": 0.7152,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1976
    },
    {
      "epoch": 0.26883328800652706,
      "grad_norm": 0.357421875,
      "learning_rate": 1.8313871745883974e-05,
      "loss": 0.6468,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1977
    },
    {
      "epoch": 0.26896926842534674,
      "grad_norm": 0.30078125,
      "learning_rate": 1.831123250927646e-05,
      "loss": 0.5311,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1978
    },
    {
      "epoch": 0.2691052488441664,
      "grad_norm": 0.609375,
      "learning_rate": 1.8308591399215523e-05,
      "loss": 0.5693,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1979
    },
    {
      "epoch": 0.2692412292629861,
      "grad_norm": 0.67578125,
      "learning_rate": 1.830594841629651e-05,
      "loss": 0.4572,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1980
    },
    {
      "epoch": 0.26937720968180584,
      "grad_norm": 0.423828125,
      "learning_rate": 1.8303303561115166e-05,
      "loss": 0.6375,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1981
    },
    {
      "epoch": 0.2695131901006255,
      "grad_norm": 0.71484375,
      "learning_rate": 1.8300656834267685e-05,
      "loss": 0.4932,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1982
    },
    {
      "epoch": 0.2696491705194452,
      "grad_norm": 0.5078125,
      "learning_rate": 1.829800823635067e-05,
      "loss": 0.7328,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1983
    },
    {
      "epoch": 0.2697851509382649,
      "grad_norm": 0.373046875,
      "learning_rate": 1.8295357767961144e-05,
      "loss": 0.4994,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1984
    },
    {
      "epoch": 0.26992113135708456,
      "grad_norm": 0.310546875,
      "learning_rate": 1.8292705429696564e-05,
      "loss": 0.5952,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1985
    },
    {
      "epoch": 0.27005711177590425,
      "grad_norm": 0.384765625,
      "learning_rate": 1.829005122215479e-05,
      "loss": 0.6763,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1986
    },
    {
      "epoch": 0.270193092194724,
      "grad_norm": 0.44921875,
      "learning_rate": 1.8287395145934114e-05,
      "loss": 0.8576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1987
    },
    {
      "epoch": 0.27032907261354366,
      "grad_norm": 0.2451171875,
      "learning_rate": 1.8284737201633248e-05,
      "loss": 0.3373,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1988
    },
    {
      "epoch": 0.27046505303236335,
      "grad_norm": 0.4375,
      "learning_rate": 1.8282077389851328e-05,
      "loss": 0.5597,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1989
    },
    {
      "epoch": 0.270601033451183,
      "grad_norm": 0.390625,
      "learning_rate": 1.8279415711187908e-05,
      "loss": 0.7062,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1990
    },
    {
      "epoch": 0.2707370138700027,
      "grad_norm": 0.73828125,
      "learning_rate": 1.827675216624296e-05,
      "loss": 0.8252,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1991
    },
    {
      "epoch": 0.2708729942888224,
      "grad_norm": 0.453125,
      "learning_rate": 1.8274086755616883e-05,
      "loss": 0.638,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1992
    },
    {
      "epoch": 0.2710089747076421,
      "grad_norm": 0.376953125,
      "learning_rate": 1.8271419479910492e-05,
      "loss": 0.5365,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1993
    },
    {
      "epoch": 0.2711449551264618,
      "grad_norm": 0.765625,
      "learning_rate": 1.8268750339725024e-05,
      "loss": 0.6348,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1994
    },
    {
      "epoch": 0.2712809355452815,
      "grad_norm": 0.330078125,
      "learning_rate": 1.8266079335662135e-05,
      "loss": 0.5845,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1995
    },
    {
      "epoch": 0.27141691596410117,
      "grad_norm": 0.38671875,
      "learning_rate": 1.8263406468323898e-05,
      "loss": 0.682,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1996
    },
    {
      "epoch": 0.27155289638292085,
      "grad_norm": 0.3984375,
      "learning_rate": 1.8260731738312817e-05,
      "loss": 0.6951,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1997
    },
    {
      "epoch": 0.27168887680174053,
      "grad_norm": 0.640625,
      "learning_rate": 1.8258055146231806e-05,
      "loss": 0.5145,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1998
    },
    {
      "epoch": 0.2718248572205602,
      "grad_norm": 0.392578125,
      "learning_rate": 1.82553766926842e-05,
      "loss": 0.5758,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 1999
    },
    {
      "epoch": 0.27196083763937995,
      "grad_norm": 0.40625,
      "learning_rate": 1.825269637827376e-05,
      "loss": 0.6991,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2000
    },
    {
      "epoch": 0.27209681805819963,
      "grad_norm": 0.63671875,
      "learning_rate": 1.8250014203604655e-05,
      "loss": 0.6502,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2001
    },
    {
      "epoch": 0.2722327984770193,
      "grad_norm": 0.3515625,
      "learning_rate": 1.8247330169281485e-05,
      "loss": 0.7607,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2002
    },
    {
      "epoch": 0.272368778895839,
      "grad_norm": 0.50390625,
      "learning_rate": 1.824464427590926e-05,
      "loss": 0.7287,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2003
    },
    {
      "epoch": 0.2725047593146587,
      "grad_norm": 0.2734375,
      "learning_rate": 1.824195652409342e-05,
      "loss": 0.5544,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2004
    },
    {
      "epoch": 0.27264073973347835,
      "grad_norm": 0.37109375,
      "learning_rate": 1.8239266914439806e-05,
      "loss": 0.6597,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2005
    },
    {
      "epoch": 0.2727767201522981,
      "grad_norm": 0.5859375,
      "learning_rate": 1.8236575447554702e-05,
      "loss": 0.7842,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2006
    },
    {
      "epoch": 0.2729127005711178,
      "grad_norm": 0.365234375,
      "learning_rate": 1.8233882124044786e-05,
      "loss": 0.7489,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2007
    },
    {
      "epoch": 0.27304868098993745,
      "grad_norm": 0.380859375,
      "learning_rate": 1.823118694451717e-05,
      "loss": 0.7355,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2008
    },
    {
      "epoch": 0.27318466140875713,
      "grad_norm": 0.5546875,
      "learning_rate": 1.8228489909579388e-05,
      "loss": 0.5291,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2009
    },
    {
      "epoch": 0.2733206418275768,
      "grad_norm": 0.50390625,
      "learning_rate": 1.8225791019839375e-05,
      "loss": 0.7109,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2010
    },
    {
      "epoch": 0.2734566222463965,
      "grad_norm": 0.52734375,
      "learning_rate": 1.8223090275905496e-05,
      "loss": 0.7505,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2011
    },
    {
      "epoch": 0.27359260266521623,
      "grad_norm": 1.4375,
      "learning_rate": 1.8220387678386536e-05,
      "loss": 0.6351,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2012
    },
    {
      "epoch": 0.2737285830840359,
      "grad_norm": 0.318359375,
      "learning_rate": 1.8217683227891687e-05,
      "loss": 0.6139,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2013
    },
    {
      "epoch": 0.2738645635028556,
      "grad_norm": 0.341796875,
      "learning_rate": 1.821497692503057e-05,
      "loss": 0.5594,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2014
    },
    {
      "epoch": 0.2740005439216753,
      "grad_norm": 0.26953125,
      "learning_rate": 1.8212268770413214e-05,
      "loss": 0.5337,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2015
    },
    {
      "epoch": 0.27413652434049496,
      "grad_norm": 0.408203125,
      "learning_rate": 1.8209558764650077e-05,
      "loss": 0.7782,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2016
    },
    {
      "epoch": 0.27427250475931464,
      "grad_norm": 0.486328125,
      "learning_rate": 1.8206846908352023e-05,
      "loss": 0.8843,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2017
    },
    {
      "epoch": 0.2744084851781344,
      "grad_norm": 0.35546875,
      "learning_rate": 1.820413320213034e-05,
      "loss": 0.5775,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2018
    },
    {
      "epoch": 0.27454446559695406,
      "grad_norm": 0.703125,
      "learning_rate": 1.8201417646596723e-05,
      "loss": 0.8452,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2019
    },
    {
      "epoch": 0.27468044601577374,
      "grad_norm": 0.7578125,
      "learning_rate": 1.81987002423633e-05,
      "loss": 0.4753,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2020
    },
    {
      "epoch": 0.2748164264345934,
      "grad_norm": 0.419921875,
      "learning_rate": 1.8195980990042607e-05,
      "loss": 0.6217,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2021
    },
    {
      "epoch": 0.2749524068534131,
      "grad_norm": 0.306640625,
      "learning_rate": 1.819325989024759e-05,
      "loss": 0.6099,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2022
    },
    {
      "epoch": 0.2750883872722328,
      "grad_norm": 0.59375,
      "learning_rate": 1.8190536943591627e-05,
      "loss": 0.764,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2023
    },
    {
      "epoch": 0.27522436769105246,
      "grad_norm": 0.41796875,
      "learning_rate": 1.8187812150688495e-05,
      "loss": 0.707,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2024
    },
    {
      "epoch": 0.2753603481098722,
      "grad_norm": 0.357421875,
      "learning_rate": 1.8185085512152402e-05,
      "loss": 0.6753,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2025
    },
    {
      "epoch": 0.2754963285286919,
      "grad_norm": 0.322265625,
      "learning_rate": 1.8182357028597963e-05,
      "loss": 0.4932,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2026
    },
    {
      "epoch": 0.27563230894751156,
      "grad_norm": 0.373046875,
      "learning_rate": 1.8179626700640208e-05,
      "loss": 0.6719,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2027
    },
    {
      "epoch": 0.27576828936633124,
      "grad_norm": 0.341796875,
      "learning_rate": 1.8176894528894595e-05,
      "loss": 0.5985,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2028
    },
    {
      "epoch": 0.2759042697851509,
      "grad_norm": 0.58203125,
      "learning_rate": 1.817416051397698e-05,
      "loss": 0.6374,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2029
    },
    {
      "epoch": 0.2760402502039706,
      "grad_norm": 0.73828125,
      "learning_rate": 1.817142465650365e-05,
      "loss": 0.5711,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2030
    },
    {
      "epoch": 0.27617623062279034,
      "grad_norm": 0.271484375,
      "learning_rate": 1.8168686957091295e-05,
      "loss": 0.474,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2031
    },
    {
      "epoch": 0.27631221104161,
      "grad_norm": 0.318359375,
      "learning_rate": 1.816594741635703e-05,
      "loss": 0.5368,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2032
    },
    {
      "epoch": 0.2764481914604297,
      "grad_norm": 0.6484375,
      "learning_rate": 1.816320603491838e-05,
      "loss": 0.7199,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2033
    },
    {
      "epoch": 0.2765841718792494,
      "grad_norm": 0.388671875,
      "learning_rate": 1.8160462813393287e-05,
      "loss": 0.6743,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2034
    },
    {
      "epoch": 0.27672015229806907,
      "grad_norm": 0.5625,
      "learning_rate": 1.8157717752400103e-05,
      "loss": 0.7402,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2035
    },
    {
      "epoch": 0.27685613271688875,
      "grad_norm": 0.375,
      "learning_rate": 1.8154970852557604e-05,
      "loss": 0.6146,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2036
    },
    {
      "epoch": 0.2769921131357085,
      "grad_norm": 0.42578125,
      "learning_rate": 1.815222211448497e-05,
      "loss": 0.6919,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2037
    },
    {
      "epoch": 0.27712809355452817,
      "grad_norm": 0.57421875,
      "learning_rate": 1.81494715388018e-05,
      "loss": 0.7627,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2038
    },
    {
      "epoch": 0.27726407397334785,
      "grad_norm": 0.50390625,
      "learning_rate": 1.814671912612811e-05,
      "loss": 0.8294,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2039
    },
    {
      "epoch": 0.2774000543921675,
      "grad_norm": 0.361328125,
      "learning_rate": 1.8143964877084327e-05,
      "loss": 0.6224,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2040
    },
    {
      "epoch": 0.2775360348109872,
      "grad_norm": 0.32421875,
      "learning_rate": 1.814120879229129e-05,
      "loss": 0.5409,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2041
    },
    {
      "epoch": 0.2776720152298069,
      "grad_norm": 0.51953125,
      "learning_rate": 1.813845087237026e-05,
      "loss": 0.8062,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2042
    },
    {
      "epoch": 0.27780799564862657,
      "grad_norm": 0.77734375,
      "learning_rate": 1.81356911179429e-05,
      "loss": 0.6432,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2043
    },
    {
      "epoch": 0.2779439760674463,
      "grad_norm": 0.484375,
      "learning_rate": 1.813292952963129e-05,
      "loss": 0.4993,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2044
    },
    {
      "epoch": 0.278079956486266,
      "grad_norm": 0.30859375,
      "learning_rate": 1.8130166108057936e-05,
      "loss": 0.5767,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2045
    },
    {
      "epoch": 0.27821593690508567,
      "grad_norm": 0.328125,
      "learning_rate": 1.812740085384574e-05,
      "loss": 0.7114,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2046
    },
    {
      "epoch": 0.27835191732390535,
      "grad_norm": 0.53515625,
      "learning_rate": 1.8124633767618024e-05,
      "loss": 0.8127,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2047
    },
    {
      "epoch": 0.27848789774272503,
      "grad_norm": 0.400390625,
      "learning_rate": 1.8121864849998526e-05,
      "loss": 0.6257,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2048
    },
    {
      "epoch": 0.2786238781615447,
      "grad_norm": 0.3515625,
      "learning_rate": 1.811909410161139e-05,
      "loss": 0.6358,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2049
    },
    {
      "epoch": 0.27875985858036445,
      "grad_norm": 0.6328125,
      "learning_rate": 1.811632152308118e-05,
      "loss": 0.8262,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2050
    },
    {
      "epoch": 0.27889583899918413,
      "grad_norm": 0.478515625,
      "learning_rate": 1.811354711503287e-05,
      "loss": 0.7145,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2051
    },
    {
      "epoch": 0.2790318194180038,
      "grad_norm": 0.474609375,
      "learning_rate": 1.811077087809184e-05,
      "loss": 0.8003,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2052
    },
    {
      "epoch": 0.2791677998368235,
      "grad_norm": 1.8359375,
      "learning_rate": 1.8107992812883894e-05,
      "loss": 0.8472,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2053
    },
    {
      "epoch": 0.2793037802556432,
      "grad_norm": 0.357421875,
      "learning_rate": 1.8105212920035238e-05,
      "loss": 0.6047,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2054
    },
    {
      "epoch": 0.27943976067446286,
      "grad_norm": 0.40234375,
      "learning_rate": 1.8102431200172496e-05,
      "loss": 0.6095,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2055
    },
    {
      "epoch": 0.2795757410932826,
      "grad_norm": 0.5078125,
      "learning_rate": 1.8099647653922697e-05,
      "loss": 0.7889,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2056
    },
    {
      "epoch": 0.2797117215121023,
      "grad_norm": 0.4921875,
      "learning_rate": 1.8096862281913292e-05,
      "loss": 0.9289,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2057
    },
    {
      "epoch": 0.27984770193092195,
      "grad_norm": 0.69140625,
      "learning_rate": 1.8094075084772137e-05,
      "loss": 0.7196,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2058
    },
    {
      "epoch": 0.27998368234974164,
      "grad_norm": 1.1484375,
      "learning_rate": 1.8091286063127498e-05,
      "loss": 0.6086,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2059
    },
    {
      "epoch": 0.2801196627685613,
      "grad_norm": 0.359375,
      "learning_rate": 1.8088495217608054e-05,
      "loss": 0.6092,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2060
    },
    {
      "epoch": 0.280255643187381,
      "grad_norm": 0.3984375,
      "learning_rate": 1.8085702548842897e-05,
      "loss": 0.5687,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2061
    },
    {
      "epoch": 0.2803916236062007,
      "grad_norm": 0.54296875,
      "learning_rate": 1.8082908057461534e-05,
      "loss": 0.276,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2062
    },
    {
      "epoch": 0.2805276040250204,
      "grad_norm": 0.2890625,
      "learning_rate": 1.8080111744093867e-05,
      "loss": 0.639,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2063
    },
    {
      "epoch": 0.2806635844438401,
      "grad_norm": 0.81640625,
      "learning_rate": 1.807731360937023e-05,
      "loss": 0.6637,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2064
    },
    {
      "epoch": 0.2807995648626598,
      "grad_norm": 0.2890625,
      "learning_rate": 1.8074513653921348e-05,
      "loss": 0.4574,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2065
    },
    {
      "epoch": 0.28093554528147946,
      "grad_norm": 0.361328125,
      "learning_rate": 1.807171187837837e-05,
      "loss": 0.6948,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2066
    },
    {
      "epoch": 0.28107152570029914,
      "grad_norm": 0.376953125,
      "learning_rate": 1.8068908283372847e-05,
      "loss": 0.6283,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2067
    },
    {
      "epoch": 0.2812075061191188,
      "grad_norm": 0.373046875,
      "learning_rate": 1.8066102869536748e-05,
      "loss": 0.7354,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2068
    },
    {
      "epoch": 0.28134348653793856,
      "grad_norm": 0.39453125,
      "learning_rate": 1.8063295637502444e-05,
      "loss": 0.6662,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2069
    },
    {
      "epoch": 0.28147946695675824,
      "grad_norm": 0.27734375,
      "learning_rate": 1.8060486587902725e-05,
      "loss": 0.5116,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2070
    },
    {
      "epoch": 0.2816154473755779,
      "grad_norm": 0.32421875,
      "learning_rate": 1.8057675721370775e-05,
      "loss": 0.6401,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2071
    },
    {
      "epoch": 0.2817514277943976,
      "grad_norm": 0.33203125,
      "learning_rate": 1.8054863038540207e-05,
      "loss": 0.5527,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2072
    },
    {
      "epoch": 0.2818874082132173,
      "grad_norm": 0.65234375,
      "learning_rate": 1.805204854004503e-05,
      "loss": 0.7959,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2073
    },
    {
      "epoch": 0.28202338863203696,
      "grad_norm": 0.431640625,
      "learning_rate": 1.8049232226519667e-05,
      "loss": 0.8118,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2074
    },
    {
      "epoch": 0.2821593690508567,
      "grad_norm": 0.318359375,
      "learning_rate": 1.8046414098598947e-05,
      "loss": 0.5879,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2075
    },
    {
      "epoch": 0.2822953494696764,
      "grad_norm": 0.349609375,
      "learning_rate": 1.8043594156918117e-05,
      "loss": 0.6038,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2076
    },
    {
      "epoch": 0.28243132988849606,
      "grad_norm": 0.359375,
      "learning_rate": 1.8040772402112824e-05,
      "loss": 0.6121,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2077
    },
    {
      "epoch": 0.28256731030731574,
      "grad_norm": 0.63671875,
      "learning_rate": 1.8037948834819122e-05,
      "loss": 0.577,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2078
    },
    {
      "epoch": 0.2827032907261354,
      "grad_norm": 0.392578125,
      "learning_rate": 1.803512345567348e-05,
      "loss": 0.6312,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2079
    },
    {
      "epoch": 0.2828392711449551,
      "grad_norm": 0.4140625,
      "learning_rate": 1.803229626531277e-05,
      "loss": 0.6178,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2080
    },
    {
      "epoch": 0.28297525156377484,
      "grad_norm": 0.4375,
      "learning_rate": 1.802946726437428e-05,
      "loss": 0.6112,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2081
    },
    {
      "epoch": 0.2831112319825945,
      "grad_norm": 0.609375,
      "learning_rate": 1.8026636453495703e-05,
      "loss": 0.5428,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2082
    },
    {
      "epoch": 0.2832472124014142,
      "grad_norm": 0.3671875,
      "learning_rate": 1.8023803833315133e-05,
      "loss": 0.5877,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2083
    },
    {
      "epoch": 0.2833831928202339,
      "grad_norm": 0.37890625,
      "learning_rate": 1.8020969404471078e-05,
      "loss": 0.6333,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2084
    },
    {
      "epoch": 0.28351917323905357,
      "grad_norm": 0.326171875,
      "learning_rate": 1.8018133167602453e-05,
      "loss": 0.5815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2085
    },
    {
      "epoch": 0.28365515365787325,
      "grad_norm": 0.51953125,
      "learning_rate": 1.8015295123348587e-05,
      "loss": 0.5859,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2086
    },
    {
      "epoch": 0.28379113407669293,
      "grad_norm": 0.5859375,
      "learning_rate": 1.80124552723492e-05,
      "loss": 0.6647,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2087
    },
    {
      "epoch": 0.28392711449551267,
      "grad_norm": 0.57421875,
      "learning_rate": 1.8009613615244438e-05,
      "loss": 0.7633,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2088
    },
    {
      "epoch": 0.28406309491433235,
      "grad_norm": 0.470703125,
      "learning_rate": 1.8006770152674836e-05,
      "loss": 0.4507,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2089
    },
    {
      "epoch": 0.28419907533315203,
      "grad_norm": 2.109375,
      "learning_rate": 1.8003924885281355e-05,
      "loss": 0.8883,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2090
    },
    {
      "epoch": 0.2843350557519717,
      "grad_norm": 0.357421875,
      "learning_rate": 1.8001077813705345e-05,
      "loss": 0.5682,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2091
    },
    {
      "epoch": 0.2844710361707914,
      "grad_norm": 0.384765625,
      "learning_rate": 1.7998228938588572e-05,
      "loss": 0.7088,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2092
    },
    {
      "epoch": 0.2846070165896111,
      "grad_norm": 0.3515625,
      "learning_rate": 1.7995378260573213e-05,
      "loss": 0.7197,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2093
    },
    {
      "epoch": 0.2847429970084308,
      "grad_norm": 0.515625,
      "learning_rate": 1.799252578030184e-05,
      "loss": 0.8449,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2094
    },
    {
      "epoch": 0.2848789774272505,
      "grad_norm": 0.369140625,
      "learning_rate": 1.798967149841744e-05,
      "loss": 0.6597,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2095
    },
    {
      "epoch": 0.28501495784607017,
      "grad_norm": 0.37109375,
      "learning_rate": 1.79868154155634e-05,
      "loss": 0.6527,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2096
    },
    {
      "epoch": 0.28515093826488985,
      "grad_norm": 0.54296875,
      "learning_rate": 1.798395753238352e-05,
      "loss": 0.569,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2097
    },
    {
      "epoch": 0.28528691868370953,
      "grad_norm": 0.55859375,
      "learning_rate": 1.7981097849522e-05,
      "loss": 0.7985,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2098
    },
    {
      "epoch": 0.2854228991025292,
      "grad_norm": 0.62890625,
      "learning_rate": 1.7978236367623448e-05,
      "loss": 0.8136,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2099
    },
    {
      "epoch": 0.28555887952134895,
      "grad_norm": 0.60546875,
      "learning_rate": 1.797537308733287e-05,
      "loss": 0.7601,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2100
    },
    {
      "epoch": 0.28569485994016863,
      "grad_norm": 0.4296875,
      "learning_rate": 1.79725080092957e-05,
      "loss": 0.6816,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2101
    },
    {
      "epoch": 0.2858308403589883,
      "grad_norm": 0.4453125,
      "learning_rate": 1.7969641134157747e-05,
      "loss": 0.8571,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2102
    },
    {
      "epoch": 0.285966820777808,
      "grad_norm": 0.37109375,
      "learning_rate": 1.7966772462565244e-05,
      "loss": 0.7616,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2103
    },
    {
      "epoch": 0.2861028011966277,
      "grad_norm": 0.51171875,
      "learning_rate": 1.796390199516483e-05,
      "loss": 0.8564,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2104
    },
    {
      "epoch": 0.28623878161544736,
      "grad_norm": 0.83984375,
      "learning_rate": 1.796102973260353e-05,
      "loss": 0.8057,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2105
    },
    {
      "epoch": 0.28637476203426704,
      "grad_norm": 0.6484375,
      "learning_rate": 1.7958155675528807e-05,
      "loss": 0.6765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2106
    },
    {
      "epoch": 0.2865107424530868,
      "grad_norm": 0.4296875,
      "learning_rate": 1.795527982458849e-05,
      "loss": 0.7568,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2107
    },
    {
      "epoch": 0.28664672287190646,
      "grad_norm": 0.400390625,
      "learning_rate": 1.7952402180430842e-05,
      "loss": 0.6777,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2108
    },
    {
      "epoch": 0.28678270329072614,
      "grad_norm": 0.32421875,
      "learning_rate": 1.7949522743704514e-05,
      "loss": 0.5923,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2109
    },
    {
      "epoch": 0.2869186837095458,
      "grad_norm": 0.279296875,
      "learning_rate": 1.794664151505857e-05,
      "loss": 0.5031,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2110
    },
    {
      "epoch": 0.2870546641283655,
      "grad_norm": 0.50390625,
      "learning_rate": 1.7943758495142472e-05,
      "loss": 0.8381,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2111
    },
    {
      "epoch": 0.2871906445471852,
      "grad_norm": 0.50390625,
      "learning_rate": 1.7940873684606085e-05,
      "loss": 0.728,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2112
    },
    {
      "epoch": 0.2873266249660049,
      "grad_norm": 0.5703125,
      "learning_rate": 1.793798708409969e-05,
      "loss": 0.7106,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2113
    },
    {
      "epoch": 0.2874626053848246,
      "grad_norm": 2.359375,
      "learning_rate": 1.793509869427395e-05,
      "loss": 0.695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2114
    },
    {
      "epoch": 0.2875985858036443,
      "grad_norm": 0.1865234375,
      "learning_rate": 1.793220851577995e-05,
      "loss": 0.3076,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2115
    },
    {
      "epoch": 0.28773456622246396,
      "grad_norm": 0.62890625,
      "learning_rate": 1.7929316549269172e-05,
      "loss": 0.5889,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2116
    },
    {
      "epoch": 0.28787054664128364,
      "grad_norm": 0.322265625,
      "learning_rate": 1.7926422795393497e-05,
      "loss": 0.6318,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2117
    },
    {
      "epoch": 0.2880065270601033,
      "grad_norm": 0.59375,
      "learning_rate": 1.7923527254805216e-05,
      "loss": 0.7321,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2118
    },
    {
      "epoch": 0.28814250747892306,
      "grad_norm": 1.1171875,
      "learning_rate": 1.792062992815702e-05,
      "loss": 0.6712,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2119
    },
    {
      "epoch": 0.28827848789774274,
      "grad_norm": 0.314453125,
      "learning_rate": 1.7917730816102003e-05,
      "loss": 0.5985,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2120
    },
    {
      "epoch": 0.2884144683165624,
      "grad_norm": 0.416015625,
      "learning_rate": 1.7914829919293653e-05,
      "loss": 0.7243,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2121
    },
    {
      "epoch": 0.2885504487353821,
      "grad_norm": 0.4609375,
      "learning_rate": 1.791192723838587e-05,
      "loss": 0.7331,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2122
    },
    {
      "epoch": 0.2886864291542018,
      "grad_norm": 0.34375,
      "learning_rate": 1.790902277403296e-05,
      "loss": 0.7521,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2123
    },
    {
      "epoch": 0.28882240957302147,
      "grad_norm": 0.69140625,
      "learning_rate": 1.7906116526889616e-05,
      "loss": 0.7485,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2124
    },
    {
      "epoch": 0.2889583899918412,
      "grad_norm": 0.3828125,
      "learning_rate": 1.790320849761095e-05,
      "loss": 0.5706,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2125
    },
    {
      "epoch": 0.2890943704106609,
      "grad_norm": 0.45703125,
      "learning_rate": 1.790029868685246e-05,
      "loss": 0.7194,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2126
    },
    {
      "epoch": 0.28923035082948056,
      "grad_norm": 0.5078125,
      "learning_rate": 1.7897387095270058e-05,
      "loss": 0.6489,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2127
    },
    {
      "epoch": 0.28936633124830025,
      "grad_norm": 0.36328125,
      "learning_rate": 1.789447372352005e-05,
      "loss": 0.7067,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2128
    },
    {
      "epoch": 0.2895023116671199,
      "grad_norm": 0.302734375,
      "learning_rate": 1.7891558572259147e-05,
      "loss": 0.6237,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2129
    },
    {
      "epoch": 0.2896382920859396,
      "grad_norm": 0.3046875,
      "learning_rate": 1.7888641642144458e-05,
      "loss": 0.592,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2130
    },
    {
      "epoch": 0.2897742725047593,
      "grad_norm": 0.5625,
      "learning_rate": 1.7885722933833496e-05,
      "loss": 0.5734,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2131
    },
    {
      "epoch": 0.289910252923579,
      "grad_norm": 0.84375,
      "learning_rate": 1.7882802447984175e-05,
      "loss": 0.7729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2132
    },
    {
      "epoch": 0.2900462333423987,
      "grad_norm": 0.314453125,
      "learning_rate": 1.7879880185254806e-05,
      "loss": 0.5316,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2133
    },
    {
      "epoch": 0.2901822137612184,
      "grad_norm": 0.3828125,
      "learning_rate": 1.78769561463041e-05,
      "loss": 0.5088,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2134
    },
    {
      "epoch": 0.29031819418003807,
      "grad_norm": 0.85546875,
      "learning_rate": 1.7874030331791178e-05,
      "loss": 0.8003,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2135
    },
    {
      "epoch": 0.29045417459885775,
      "grad_norm": 0.349609375,
      "learning_rate": 1.787110274237555e-05,
      "loss": 0.5327,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2136
    },
    {
      "epoch": 0.29059015501767743,
      "grad_norm": 0.4453125,
      "learning_rate": 1.786817337871713e-05,
      "loss": 0.688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2137
    },
    {
      "epoch": 0.29072613543649717,
      "grad_norm": 0.2890625,
      "learning_rate": 1.7865242241476237e-05,
      "loss": 0.5137,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2138
    },
    {
      "epoch": 0.29086211585531685,
      "grad_norm": 0.30078125,
      "learning_rate": 1.7862309331313578e-05,
      "loss": 0.5898,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2139
    },
    {
      "epoch": 0.29099809627413653,
      "grad_norm": 0.443359375,
      "learning_rate": 1.785937464889027e-05,
      "loss": 0.7388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2140
    },
    {
      "epoch": 0.2911340766929562,
      "grad_norm": 0.76171875,
      "learning_rate": 1.785643819486783e-05,
      "loss": 0.6136,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2141
    },
    {
      "epoch": 0.2912700571117759,
      "grad_norm": 0.416015625,
      "learning_rate": 1.7853499969908163e-05,
      "loss": 0.7539,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2142
    },
    {
      "epoch": 0.2914060375305956,
      "grad_norm": 0.2392578125,
      "learning_rate": 1.7850559974673588e-05,
      "loss": 0.4494,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2143
    },
    {
      "epoch": 0.2915420179494153,
      "grad_norm": 0.54296875,
      "learning_rate": 1.7847618209826813e-05,
      "loss": 0.6126,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2144
    },
    {
      "epoch": 0.291677998368235,
      "grad_norm": 1.1171875,
      "learning_rate": 1.784467467603095e-05,
      "loss": 0.8721,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2145
    },
    {
      "epoch": 0.2918139787870547,
      "grad_norm": 0.609375,
      "learning_rate": 1.78417293739495e-05,
      "loss": 0.7767,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2146
    },
    {
      "epoch": 0.29194995920587435,
      "grad_norm": 0.287109375,
      "learning_rate": 1.7838782304246378e-05,
      "loss": 0.5493,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2147
    },
    {
      "epoch": 0.29208593962469404,
      "grad_norm": 0.4296875,
      "learning_rate": 1.7835833467585887e-05,
      "loss": 0.6755,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2148
    },
    {
      "epoch": 0.2922219200435137,
      "grad_norm": 0.71875,
      "learning_rate": 1.7832882864632734e-05,
      "loss": 0.7454,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2149
    },
    {
      "epoch": 0.2923579004623334,
      "grad_norm": 0.40234375,
      "learning_rate": 1.7829930496052014e-05,
      "loss": 0.763,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2150
    },
    {
      "epoch": 0.29249388088115313,
      "grad_norm": 0.53515625,
      "learning_rate": 1.7826976362509236e-05,
      "loss": 0.6618,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2151
    },
    {
      "epoch": 0.2926298612999728,
      "grad_norm": 0.263671875,
      "learning_rate": 1.7824020464670286e-05,
      "loss": 0.5439,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2152
    },
    {
      "epoch": 0.2927658417187925,
      "grad_norm": 0.28125,
      "learning_rate": 1.782106280320147e-05,
      "loss": 0.4549,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2153
    },
    {
      "epoch": 0.2929018221376122,
      "grad_norm": 0.482421875,
      "learning_rate": 1.7818103378769474e-05,
      "loss": 0.7326,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2154
    },
    {
      "epoch": 0.29303780255643186,
      "grad_norm": 0.7734375,
      "learning_rate": 1.7815142192041393e-05,
      "loss": 0.5452,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2155
    },
    {
      "epoch": 0.29317378297525154,
      "grad_norm": 0.349609375,
      "learning_rate": 1.781217924368472e-05,
      "loss": 0.5301,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2156
    },
    {
      "epoch": 0.2933097633940713,
      "grad_norm": 0.625,
      "learning_rate": 1.7809214534367324e-05,
      "loss": 0.6291,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2157
    },
    {
      "epoch": 0.29344574381289096,
      "grad_norm": 0.33984375,
      "learning_rate": 1.78062480647575e-05,
      "loss": 0.5648,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2158
    },
    {
      "epoch": 0.29358172423171064,
      "grad_norm": 0.5234375,
      "learning_rate": 1.780327983552392e-05,
      "loss": 0.6283,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2159
    },
    {
      "epoch": 0.2937177046505303,
      "grad_norm": 0.455078125,
      "learning_rate": 1.7800309847335664e-05,
      "loss": 0.7257,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2160
    },
    {
      "epoch": 0.29385368506935,
      "grad_norm": 0.31640625,
      "learning_rate": 1.7797338100862204e-05,
      "loss": 0.5111,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2161
    },
    {
      "epoch": 0.2939896654881697,
      "grad_norm": 0.41015625,
      "learning_rate": 1.77943645967734e-05,
      "loss": 0.6553,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2162
    },
    {
      "epoch": 0.2941256459069894,
      "grad_norm": 0.52734375,
      "learning_rate": 1.779138933573952e-05,
      "loss": 0.4943,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2163
    },
    {
      "epoch": 0.2942616263258091,
      "grad_norm": 0.392578125,
      "learning_rate": 1.7788412318431232e-05,
      "loss": 0.7936,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2164
    },
    {
      "epoch": 0.2943976067446288,
      "grad_norm": 0.326171875,
      "learning_rate": 1.778543354551958e-05,
      "loss": 0.5815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2165
    },
    {
      "epoch": 0.29453358716344846,
      "grad_norm": 0.3203125,
      "learning_rate": 1.7782453017676025e-05,
      "loss": 0.5498,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2166
    },
    {
      "epoch": 0.29466956758226814,
      "grad_norm": 0.4453125,
      "learning_rate": 1.777947073557241e-05,
      "loss": 0.613,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2167
    },
    {
      "epoch": 0.2948055480010878,
      "grad_norm": 0.341796875,
      "learning_rate": 1.777648669988098e-05,
      "loss": 0.7041,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2168
    },
    {
      "epoch": 0.2949415284199075,
      "grad_norm": 0.267578125,
      "learning_rate": 1.777350091127437e-05,
      "loss": 0.5096,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2169
    },
    {
      "epoch": 0.29507750883872724,
      "grad_norm": 0.294921875,
      "learning_rate": 1.7770513370425615e-05,
      "loss": 0.5145,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2170
    },
    {
      "epoch": 0.2952134892575469,
      "grad_norm": 1.3125,
      "learning_rate": 1.7767524078008144e-05,
      "loss": 0.9092,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2171
    },
    {
      "epoch": 0.2953494696763666,
      "grad_norm": 0.59375,
      "learning_rate": 1.776453303469578e-05,
      "loss": 0.6903,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2172
    },
    {
      "epoch": 0.2954854500951863,
      "grad_norm": 0.74609375,
      "learning_rate": 1.776154024116274e-05,
      "loss": 0.7041,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2173
    },
    {
      "epoch": 0.29562143051400597,
      "grad_norm": 0.310546875,
      "learning_rate": 1.7758545698083636e-05,
      "loss": 0.5462,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2174
    },
    {
      "epoch": 0.29575741093282565,
      "grad_norm": 0.41015625,
      "learning_rate": 1.7755549406133474e-05,
      "loss": 0.6986,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2175
    },
    {
      "epoch": 0.2958933913516454,
      "grad_norm": 0.578125,
      "learning_rate": 1.775255136598766e-05,
      "loss": 0.7378,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2176
    },
    {
      "epoch": 0.29602937177046507,
      "grad_norm": 0.65234375,
      "learning_rate": 1.774955157832198e-05,
      "loss": 0.5692,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2177
    },
    {
      "epoch": 0.29616535218928475,
      "grad_norm": 0.375,
      "learning_rate": 1.7746550043812627e-05,
      "loss": 0.5942,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2178
    },
    {
      "epoch": 0.29630133260810443,
      "grad_norm": 0.296875,
      "learning_rate": 1.7743546763136187e-05,
      "loss": 0.4686,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2179
    },
    {
      "epoch": 0.2964373130269241,
      "grad_norm": 0.34375,
      "learning_rate": 1.7740541736969635e-05,
      "loss": 0.4427,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2180
    },
    {
      "epoch": 0.2965732934457438,
      "grad_norm": 0.2890625,
      "learning_rate": 1.7737534965990336e-05,
      "loss": 0.5457,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2181
    },
    {
      "epoch": 0.2967092738645635,
      "grad_norm": 0.61328125,
      "learning_rate": 1.7734526450876057e-05,
      "loss": 0.8576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2182
    },
    {
      "epoch": 0.2968452542833832,
      "grad_norm": 0.50390625,
      "learning_rate": 1.7731516192304955e-05,
      "loss": 0.8071,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2183
    },
    {
      "epoch": 0.2969812347022029,
      "grad_norm": 0.63671875,
      "learning_rate": 1.7728504190955574e-05,
      "loss": 0.6466,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2184
    },
    {
      "epoch": 0.29711721512102257,
      "grad_norm": 0.40625,
      "learning_rate": 1.772549044750686e-05,
      "loss": 0.5923,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2185
    },
    {
      "epoch": 0.29725319553984225,
      "grad_norm": 0.40625,
      "learning_rate": 1.772247496263815e-05,
      "loss": 0.5275,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2186
    },
    {
      "epoch": 0.29738917595866193,
      "grad_norm": 0.37890625,
      "learning_rate": 1.7719457737029162e-05,
      "loss": 0.6652,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2187
    },
    {
      "epoch": 0.29752515637748167,
      "grad_norm": 0.6328125,
      "learning_rate": 1.7716438771360023e-05,
      "loss": 0.7243,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2188
    },
    {
      "epoch": 0.29766113679630135,
      "grad_norm": 0.34375,
      "learning_rate": 1.7713418066311243e-05,
      "loss": 0.6045,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2189
    },
    {
      "epoch": 0.29779711721512103,
      "grad_norm": 0.3671875,
      "learning_rate": 1.7710395622563726e-05,
      "loss": 0.6453,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2190
    },
    {
      "epoch": 0.2979330976339407,
      "grad_norm": 0.453125,
      "learning_rate": 1.770737144079877e-05,
      "loss": 0.6636,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2191
    },
    {
      "epoch": 0.2980690780527604,
      "grad_norm": 24.375,
      "learning_rate": 1.7704345521698057e-05,
      "loss": 0.9497,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2192
    },
    {
      "epoch": 0.2982050584715801,
      "grad_norm": 0.5859375,
      "learning_rate": 1.7701317865943675e-05,
      "loss": 0.9126,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2193
    },
    {
      "epoch": 0.29834103889039976,
      "grad_norm": 0.41015625,
      "learning_rate": 1.7698288474218087e-05,
      "loss": 0.6725,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2194
    },
    {
      "epoch": 0.2984770193092195,
      "grad_norm": 0.283203125,
      "learning_rate": 1.7695257347204157e-05,
      "loss": 0.6405,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2195
    },
    {
      "epoch": 0.2986129997280392,
      "grad_norm": 0.40625,
      "learning_rate": 1.7692224485585136e-05,
      "loss": 0.5379,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2196
    },
    {
      "epoch": 0.29874898014685886,
      "grad_norm": 0.44140625,
      "learning_rate": 1.768918989004467e-05,
      "loss": 0.5996,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2197
    },
    {
      "epoch": 0.29888496056567854,
      "grad_norm": 0.875,
      "learning_rate": 1.7686153561266794e-05,
      "loss": 0.7733,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2198
    },
    {
      "epoch": 0.2990209409844982,
      "grad_norm": 0.53515625,
      "learning_rate": 1.7683115499935936e-05,
      "loss": 0.6276,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2199
    },
    {
      "epoch": 0.2991569214033179,
      "grad_norm": 0.3125,
      "learning_rate": 1.768007570673691e-05,
      "loss": 0.6169,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2200
    },
    {
      "epoch": 0.29929290182213764,
      "grad_norm": 0.62890625,
      "learning_rate": 1.767703418235492e-05,
      "loss": 0.7487,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2201
    },
    {
      "epoch": 0.2994288822409573,
      "grad_norm": 0.466796875,
      "learning_rate": 1.7673990927475564e-05,
      "loss": 0.6548,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2202
    },
    {
      "epoch": 0.299564862659777,
      "grad_norm": 0.890625,
      "learning_rate": 1.7670945942784835e-05,
      "loss": 0.6946,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2203
    },
    {
      "epoch": 0.2997008430785967,
      "grad_norm": 0.421875,
      "learning_rate": 1.7667899228969104e-05,
      "loss": 0.6466,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2204
    },
    {
      "epoch": 0.29983682349741636,
      "grad_norm": 0.359375,
      "learning_rate": 1.766485078671514e-05,
      "loss": 0.5726,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2205
    },
    {
      "epoch": 0.29997280391623604,
      "grad_norm": 0.298828125,
      "learning_rate": 1.766180061671009e-05,
      "loss": 0.5871,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2206
    },
    {
      "epoch": 0.3001087843350558,
      "grad_norm": 0.306640625,
      "learning_rate": 1.7658748719641515e-05,
      "loss": 0.5509,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2207
    },
    {
      "epoch": 0.30024476475387546,
      "grad_norm": 0.412109375,
      "learning_rate": 1.765569509619734e-05,
      "loss": 0.68,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2208
    },
    {
      "epoch": 0.30038074517269514,
      "grad_norm": 0.7578125,
      "learning_rate": 1.7652639747065886e-05,
      "loss": 0.5027,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2209
    },
    {
      "epoch": 0.3005167255915148,
      "grad_norm": 0.453125,
      "learning_rate": 1.7649582672935878e-05,
      "loss": 0.6522,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2210
    },
    {
      "epoch": 0.3006527060103345,
      "grad_norm": 0.7890625,
      "learning_rate": 1.764652387449641e-05,
      "loss": 0.5099,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2211
    },
    {
      "epoch": 0.3007886864291542,
      "grad_norm": 0.40625,
      "learning_rate": 1.764346335243697e-05,
      "loss": 0.6177,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2212
    },
    {
      "epoch": 0.30092466684797387,
      "grad_norm": 0.6171875,
      "learning_rate": 1.764040110744744e-05,
      "loss": 0.7062,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2213
    },
    {
      "epoch": 0.3010606472667936,
      "grad_norm": 0.41796875,
      "learning_rate": 1.7637337140218094e-05,
      "loss": 0.7668,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2214
    },
    {
      "epoch": 0.3011966276856133,
      "grad_norm": 0.5546875,
      "learning_rate": 1.763427145143958e-05,
      "loss": 0.9679,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2215
    },
    {
      "epoch": 0.30133260810443296,
      "grad_norm": 0.376953125,
      "learning_rate": 1.7631204041802942e-05,
      "loss": 0.6302,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2216
    },
    {
      "epoch": 0.30146858852325265,
      "grad_norm": 0.4453125,
      "learning_rate": 1.7628134911999613e-05,
      "loss": 0.6004,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2217
    },
    {
      "epoch": 0.3016045689420723,
      "grad_norm": 0.283203125,
      "learning_rate": 1.7625064062721414e-05,
      "loss": 0.5106,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2218
    },
    {
      "epoch": 0.301740549360892,
      "grad_norm": 0.326171875,
      "learning_rate": 1.762199149466055e-05,
      "loss": 0.5798,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2219
    },
    {
      "epoch": 0.30187652977971174,
      "grad_norm": 10.1875,
      "learning_rate": 1.7618917208509616e-05,
      "loss": 0.8766,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2220
    },
    {
      "epoch": 0.3020125101985314,
      "grad_norm": 0.291015625,
      "learning_rate": 1.7615841204961594e-05,
      "loss": 0.4756,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2221
    },
    {
      "epoch": 0.3021484906173511,
      "grad_norm": 0.43359375,
      "learning_rate": 1.7612763484709855e-05,
      "loss": 0.6502,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2222
    },
    {
      "epoch": 0.3022844710361708,
      "grad_norm": 0.478515625,
      "learning_rate": 1.7609684048448148e-05,
      "loss": 0.7244,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2223
    },
    {
      "epoch": 0.30242045145499047,
      "grad_norm": 0.6171875,
      "learning_rate": 1.7606602896870625e-05,
      "loss": 0.9297,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2224
    },
    {
      "epoch": 0.30255643187381015,
      "grad_norm": 0.314453125,
      "learning_rate": 1.7603520030671805e-05,
      "loss": 0.6444,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2225
    },
    {
      "epoch": 0.3026924122926299,
      "grad_norm": 0.34765625,
      "learning_rate": 1.760043545054661e-05,
      "loss": 0.7231,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2226
    },
    {
      "epoch": 0.30282839271144957,
      "grad_norm": 0.369140625,
      "learning_rate": 1.759734915719034e-05,
      "loss": 0.579,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2227
    },
    {
      "epoch": 0.30296437313026925,
      "grad_norm": 0.86328125,
      "learning_rate": 1.7594261151298687e-05,
      "loss": 0.9325,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2228
    },
    {
      "epoch": 0.30310035354908893,
      "grad_norm": 0.3125,
      "learning_rate": 1.759117143356772e-05,
      "loss": 0.5729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2229
    },
    {
      "epoch": 0.3032363339679086,
      "grad_norm": 0.40625,
      "learning_rate": 1.7588080004693905e-05,
      "loss": 0.6753,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2230
    },
    {
      "epoch": 0.3033723143867283,
      "grad_norm": 0.46484375,
      "learning_rate": 1.7584986865374084e-05,
      "loss": 0.7067,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2231
    },
    {
      "epoch": 0.303508294805548,
      "grad_norm": 0.412109375,
      "learning_rate": 1.7581892016305486e-05,
      "loss": 0.6842,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2232
    },
    {
      "epoch": 0.3036442752243677,
      "grad_norm": 0.302734375,
      "learning_rate": 1.757879545818573e-05,
      "loss": 0.571,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2233
    },
    {
      "epoch": 0.3037802556431874,
      "grad_norm": 0.53515625,
      "learning_rate": 1.7575697191712823e-05,
      "loss": 0.519,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2234
    },
    {
      "epoch": 0.3039162360620071,
      "grad_norm": 0.44140625,
      "learning_rate": 1.7572597217585146e-05,
      "loss": 0.6558,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2235
    },
    {
      "epoch": 0.30405221648082675,
      "grad_norm": 0.376953125,
      "learning_rate": 1.756949553650147e-05,
      "loss": 0.8545,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2236
    },
    {
      "epoch": 0.30418819689964643,
      "grad_norm": 0.40625,
      "learning_rate": 1.756639214916096e-05,
      "loss": 0.5576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2237
    },
    {
      "epoch": 0.3043241773184661,
      "grad_norm": 0.3359375,
      "learning_rate": 1.7563287056263147e-05,
      "loss": 0.5124,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2238
    },
    {
      "epoch": 0.30446015773728585,
      "grad_norm": 0.29296875,
      "learning_rate": 1.7560180258507968e-05,
      "loss": 0.529,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2239
    },
    {
      "epoch": 0.30459613815610553,
      "grad_norm": 0.53125,
      "learning_rate": 1.7557071756595723e-05,
      "loss": 0.6012,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2240
    },
    {
      "epoch": 0.3047321185749252,
      "grad_norm": 0.390625,
      "learning_rate": 1.7553961551227112e-05,
      "loss": 0.6462,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2241
    },
    {
      "epoch": 0.3048680989937449,
      "grad_norm": 0.427734375,
      "learning_rate": 1.755084964310321e-05,
      "loss": 0.8926,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2242
    },
    {
      "epoch": 0.3050040794125646,
      "grad_norm": 0.373046875,
      "learning_rate": 1.7547736032925485e-05,
      "loss": 0.7407,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2243
    },
    {
      "epoch": 0.30514005983138426,
      "grad_norm": 0.349609375,
      "learning_rate": 1.754462072139578e-05,
      "loss": 0.6595,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2244
    },
    {
      "epoch": 0.305276040250204,
      "grad_norm": 0.54296875,
      "learning_rate": 1.754150370921632e-05,
      "loss": 0.8905,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2245
    },
    {
      "epoch": 0.3054120206690237,
      "grad_norm": 1.140625,
      "learning_rate": 1.753838499708972e-05,
      "loss": 0.6564,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2246
    },
    {
      "epoch": 0.30554800108784336,
      "grad_norm": 0.423828125,
      "learning_rate": 1.753526458571898e-05,
      "loss": 0.5496,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2247
    },
    {
      "epoch": 0.30568398150666304,
      "grad_norm": 0.58203125,
      "learning_rate": 1.753214247580747e-05,
      "loss": 0.4417,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2248
    },
    {
      "epoch": 0.3058199619254827,
      "grad_norm": 0.625,
      "learning_rate": 1.7529018668058962e-05,
      "loss": 0.6899,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2249
    },
    {
      "epoch": 0.3059559423443024,
      "grad_norm": 0.453125,
      "learning_rate": 1.7525893163177596e-05,
      "loss": 0.7827,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2250
    },
    {
      "epoch": 0.30609192276312214,
      "grad_norm": 0.765625,
      "learning_rate": 1.7522765961867896e-05,
      "loss": 0.7653,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2251
    },
    {
      "epoch": 0.3062279031819418,
      "grad_norm": 0.326171875,
      "learning_rate": 1.7519637064834777e-05,
      "loss": 0.5096,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2252
    },
    {
      "epoch": 0.3063638836007615,
      "grad_norm": 0.447265625,
      "learning_rate": 1.7516506472783525e-05,
      "loss": 0.5884,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2253
    },
    {
      "epoch": 0.3064998640195812,
      "grad_norm": 0.5234375,
      "learning_rate": 1.7513374186419822e-05,
      "loss": 0.8604,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2254
    },
    {
      "epoch": 0.30663584443840086,
      "grad_norm": 0.3359375,
      "learning_rate": 1.7510240206449712e-05,
      "loss": 0.5282,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2255
    },
    {
      "epoch": 0.30677182485722054,
      "grad_norm": 0.6953125,
      "learning_rate": 1.750710453357964e-05,
      "loss": 0.8735,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2256
    },
    {
      "epoch": 0.3069078052760402,
      "grad_norm": 0.435546875,
      "learning_rate": 1.7503967168516426e-05,
      "loss": 0.6751,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2257
    },
    {
      "epoch": 0.30704378569485996,
      "grad_norm": 0.4453125,
      "learning_rate": 1.7500828111967266e-05,
      "loss": 0.7193,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2258
    },
    {
      "epoch": 0.30717976611367964,
      "grad_norm": 0.330078125,
      "learning_rate": 1.7497687364639747e-05,
      "loss": 0.5326,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2259
    },
    {
      "epoch": 0.3073157465324993,
      "grad_norm": 0.376953125,
      "learning_rate": 1.7494544927241828e-05,
      "loss": 0.6335,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2260
    },
    {
      "epoch": 0.307451726951319,
      "grad_norm": 0.50390625,
      "learning_rate": 1.7491400800481853e-05,
      "loss": 0.7196,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2261
    },
    {
      "epoch": 0.3075877073701387,
      "grad_norm": 0.3984375,
      "learning_rate": 1.7488254985068547e-05,
      "loss": 0.6906,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2262
    },
    {
      "epoch": 0.30772368778895837,
      "grad_norm": 0.33203125,
      "learning_rate": 1.7485107481711014e-05,
      "loss": 0.4969,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2263
    },
    {
      "epoch": 0.3078596682077781,
      "grad_norm": 0.65234375,
      "learning_rate": 1.748195829111874e-05,
      "loss": 0.6624,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2264
    },
    {
      "epoch": 0.3079956486265978,
      "grad_norm": 0.5703125,
      "learning_rate": 1.7478807414001595e-05,
      "loss": 0.8815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2265
    },
    {
      "epoch": 0.30813162904541747,
      "grad_norm": 0.396484375,
      "learning_rate": 1.7475654851069822e-05,
      "loss": 0.659,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2266
    },
    {
      "epoch": 0.30826760946423715,
      "grad_norm": 0.275390625,
      "learning_rate": 1.7472500603034047e-05,
      "loss": 0.5151,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2267
    },
    {
      "epoch": 0.30840358988305683,
      "grad_norm": 0.3203125,
      "learning_rate": 1.7469344670605274e-05,
      "loss": 0.5885,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2268
    },
    {
      "epoch": 0.3085395703018765,
      "grad_norm": 0.349609375,
      "learning_rate": 1.7466187054494895e-05,
      "loss": 0.5451,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2269
    },
    {
      "epoch": 0.30867555072069625,
      "grad_norm": 0.4609375,
      "learning_rate": 1.746302775541467e-05,
      "loss": 0.4784,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2270
    },
    {
      "epoch": 0.3088115311395159,
      "grad_norm": 0.388671875,
      "learning_rate": 1.7459866774076748e-05,
      "loss": 0.6799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2271
    },
    {
      "epoch": 0.3089475115583356,
      "grad_norm": 0.640625,
      "learning_rate": 1.745670411119365e-05,
      "loss": 0.5946,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2272
    },
    {
      "epoch": 0.3090834919771553,
      "grad_norm": 0.6328125,
      "learning_rate": 1.7453539767478276e-05,
      "loss": 0.7637,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2273
    },
    {
      "epoch": 0.30921947239597497,
      "grad_norm": 0.671875,
      "learning_rate": 1.745037374364391e-05,
      "loss": 0.8657,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2274
    },
    {
      "epoch": 0.30935545281479465,
      "grad_norm": 0.2353515625,
      "learning_rate": 1.7447206040404218e-05,
      "loss": 0.4271,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2275
    },
    {
      "epoch": 0.30949143323361433,
      "grad_norm": 0.5625,
      "learning_rate": 1.744403665847323e-05,
      "loss": 0.7925,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2276
    },
    {
      "epoch": 0.30962741365243407,
      "grad_norm": 0.25,
      "learning_rate": 1.7440865598565372e-05,
      "loss": 0.5208,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2277
    },
    {
      "epoch": 0.30976339407125375,
      "grad_norm": 0.41796875,
      "learning_rate": 1.7437692861395433e-05,
      "loss": 0.6481,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2278
    },
    {
      "epoch": 0.30989937449007343,
      "grad_norm": 0.2353515625,
      "learning_rate": 1.7434518447678586e-05,
      "loss": 0.4157,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2279
    },
    {
      "epoch": 0.3100353549088931,
      "grad_norm": 0.423828125,
      "learning_rate": 1.7431342358130394e-05,
      "loss": 0.6346,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2280
    },
    {
      "epoch": 0.3101713353277128,
      "grad_norm": 0.84765625,
      "learning_rate": 1.7428164593466772e-05,
      "loss": 0.6688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2281
    },
    {
      "epoch": 0.3103073157465325,
      "grad_norm": 0.5390625,
      "learning_rate": 1.7424985154404036e-05,
      "loss": 0.6909,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2282
    },
    {
      "epoch": 0.3104432961653522,
      "grad_norm": 0.32421875,
      "learning_rate": 1.7421804041658867e-05,
      "loss": 0.493,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2283
    },
    {
      "epoch": 0.3105792765841719,
      "grad_norm": 0.326171875,
      "learning_rate": 1.7418621255948325e-05,
      "loss": 0.4271,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2284
    },
    {
      "epoch": 0.3107152570029916,
      "grad_norm": 1.296875,
      "learning_rate": 1.7415436797989856e-05,
      "loss": 0.7603,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2285
    },
    {
      "epoch": 0.31085123742181126,
      "grad_norm": 0.859375,
      "learning_rate": 1.7412250668501268e-05,
      "loss": 0.6764,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2286
    },
    {
      "epoch": 0.31098721784063094,
      "grad_norm": 0.3828125,
      "learning_rate": 1.7409062868200756e-05,
      "loss": 0.7093,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2287
    },
    {
      "epoch": 0.3111231982594506,
      "grad_norm": 0.375,
      "learning_rate": 1.740587339780689e-05,
      "loss": 0.5951,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2288
    },
    {
      "epoch": 0.31125917867827035,
      "grad_norm": 0.392578125,
      "learning_rate": 1.7402682258038615e-05,
      "loss": 0.6222,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2289
    },
    {
      "epoch": 0.31139515909709004,
      "grad_norm": 0.283203125,
      "learning_rate": 1.7399489449615258e-05,
      "loss": 0.5485,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2290
    },
    {
      "epoch": 0.3115311395159097,
      "grad_norm": 0.466796875,
      "learning_rate": 1.7396294973256508e-05,
      "loss": 0.8643,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2291
    },
    {
      "epoch": 0.3116671199347294,
      "grad_norm": 0.3828125,
      "learning_rate": 1.7393098829682445e-05,
      "loss": 0.6074,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2292
    },
    {
      "epoch": 0.3118031003535491,
      "grad_norm": 0.337890625,
      "learning_rate": 1.7389901019613512e-05,
      "loss": 0.439,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2293
    },
    {
      "epoch": 0.31193908077236876,
      "grad_norm": 0.359375,
      "learning_rate": 1.7386701543770544e-05,
      "loss": 0.5749,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2294
    },
    {
      "epoch": 0.3120750611911885,
      "grad_norm": 0.4140625,
      "learning_rate": 1.738350040287474e-05,
      "loss": 0.6035,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2295
    },
    {
      "epoch": 0.3122110416100082,
      "grad_norm": 0.423828125,
      "learning_rate": 1.738029759764767e-05,
      "loss": 0.694,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2296
    },
    {
      "epoch": 0.31234702202882786,
      "grad_norm": 0.48046875,
      "learning_rate": 1.737709312881129e-05,
      "loss": 0.7967,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2297
    },
    {
      "epoch": 0.31248300244764754,
      "grad_norm": 0.482421875,
      "learning_rate": 1.7373886997087925e-05,
      "loss": 0.7596,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2298
    },
    {
      "epoch": 0.3126189828664672,
      "grad_norm": 0.359375,
      "learning_rate": 1.737067920320028e-05,
      "loss": 0.6395,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2299
    },
    {
      "epoch": 0.3127549632852869,
      "grad_norm": 0.50390625,
      "learning_rate": 1.7367469747871428e-05,
      "loss": 0.7949,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2300
    },
    {
      "epoch": 0.3128909437041066,
      "grad_norm": 0.263671875,
      "learning_rate": 1.7364258631824817e-05,
      "loss": 0.5312,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2301
    },
    {
      "epoch": 0.3130269241229263,
      "grad_norm": 0.3359375,
      "learning_rate": 1.7361045855784277e-05,
      "loss": 0.5137,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2302
    },
    {
      "epoch": 0.313162904541746,
      "grad_norm": 0.291015625,
      "learning_rate": 1.7357831420474002e-05,
      "loss": 0.5044,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2303
    },
    {
      "epoch": 0.3132988849605657,
      "grad_norm": 0.55859375,
      "learning_rate": 1.735461532661857e-05,
      "loss": 0.6123,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2304
    },
    {
      "epoch": 0.31343486537938536,
      "grad_norm": 0.314453125,
      "learning_rate": 1.735139757494292e-05,
      "loss": 0.3621,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2305
    },
    {
      "epoch": 0.31357084579820504,
      "grad_norm": 0.60546875,
      "learning_rate": 1.7348178166172385e-05,
      "loss": 0.6084,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2306
    },
    {
      "epoch": 0.3137068262170247,
      "grad_norm": 0.458984375,
      "learning_rate": 1.7344957101032646e-05,
      "loss": 0.842,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2307
    },
    {
      "epoch": 0.31384280663584446,
      "grad_norm": 0.5546875,
      "learning_rate": 1.734173438024978e-05,
      "loss": 0.5597,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2308
    },
    {
      "epoch": 0.31397878705466414,
      "grad_norm": 0.3046875,
      "learning_rate": 1.7338510004550225e-05,
      "loss": 0.6029,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2309
    },
    {
      "epoch": 0.3141147674734838,
      "grad_norm": 0.244140625,
      "learning_rate": 1.7335283974660792e-05,
      "loss": 0.4443,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2310
    },
    {
      "epoch": 0.3142507478923035,
      "grad_norm": 0.30859375,
      "learning_rate": 1.7332056291308667e-05,
      "loss": 0.4058,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2311
    },
    {
      "epoch": 0.3143867283111232,
      "grad_norm": 0.30859375,
      "learning_rate": 1.7328826955221413e-05,
      "loss": 0.5728,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2312
    },
    {
      "epoch": 0.31452270872994287,
      "grad_norm": 0.375,
      "learning_rate": 1.732559596712696e-05,
      "loss": 0.7783,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2313
    },
    {
      "epoch": 0.3146586891487626,
      "grad_norm": 0.37890625,
      "learning_rate": 1.7322363327753616e-05,
      "loss": 0.6107,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2314
    },
    {
      "epoch": 0.3147946695675823,
      "grad_norm": 0.361328125,
      "learning_rate": 1.7319129037830054e-05,
      "loss": 0.5576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2315
    },
    {
      "epoch": 0.31493064998640197,
      "grad_norm": 0.375,
      "learning_rate": 1.731589309808532e-05,
      "loss": 0.6574,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2316
    },
    {
      "epoch": 0.31506663040522165,
      "grad_norm": 0.333984375,
      "learning_rate": 1.7312655509248837e-05,
      "loss": 0.6323,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2317
    },
    {
      "epoch": 0.31520261082404133,
      "grad_norm": 0.427734375,
      "learning_rate": 1.7309416272050395e-05,
      "loss": 0.5799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2318
    },
    {
      "epoch": 0.315338591242861,
      "grad_norm": 0.427734375,
      "learning_rate": 1.7306175387220165e-05,
      "loss": 0.7847,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2319
    },
    {
      "epoch": 0.3154745716616807,
      "grad_norm": 0.484375,
      "learning_rate": 1.7302932855488675e-05,
      "loss": 0.7061,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2320
    },
    {
      "epoch": 0.31561055208050043,
      "grad_norm": 0.65625,
      "learning_rate": 1.7299688677586832e-05,
      "loss": 0.8216,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2321
    },
    {
      "epoch": 0.3157465324993201,
      "grad_norm": 0.46875,
      "learning_rate": 1.7296442854245915e-05,
      "loss": 0.5907,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2322
    },
    {
      "epoch": 0.3158825129181398,
      "grad_norm": 0.44921875,
      "learning_rate": 1.7293195386197573e-05,
      "loss": 0.6371,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2323
    },
    {
      "epoch": 0.31601849333695947,
      "grad_norm": 0.7265625,
      "learning_rate": 1.728994627417382e-05,
      "loss": 0.7026,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2324
    },
    {
      "epoch": 0.31615447375577915,
      "grad_norm": 1.546875,
      "learning_rate": 1.7286695518907052e-05,
      "loss": 0.7371,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2325
    },
    {
      "epoch": 0.31629045417459883,
      "grad_norm": 3.15625,
      "learning_rate": 1.7283443121130027e-05,
      "loss": 0.792,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2326
    },
    {
      "epoch": 0.31642643459341857,
      "grad_norm": 0.40625,
      "learning_rate": 1.7280189081575874e-05,
      "loss": 0.6868,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2327
    },
    {
      "epoch": 0.31656241501223825,
      "grad_norm": 0.345703125,
      "learning_rate": 1.727693340097809e-05,
      "loss": 0.6086,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2328
    },
    {
      "epoch": 0.31669839543105793,
      "grad_norm": 0.3359375,
      "learning_rate": 1.7273676080070554e-05,
      "loss": 0.5793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2329
    },
    {
      "epoch": 0.3168343758498776,
      "grad_norm": 0.322265625,
      "learning_rate": 1.7270417119587503e-05,
      "loss": 0.5441,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2330
    },
    {
      "epoch": 0.3169703562686973,
      "grad_norm": 1.359375,
      "learning_rate": 1.726715652026354e-05,
      "loss": 0.9119,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2331
    },
    {
      "epoch": 0.317106336687517,
      "grad_norm": 0.34765625,
      "learning_rate": 1.726389428283365e-05,
      "loss": 0.6991,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2332
    },
    {
      "epoch": 0.3172423171063367,
      "grad_norm": 0.703125,
      "learning_rate": 1.7260630408033176e-05,
      "loss": 0.8462,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2333
    },
    {
      "epoch": 0.3173782975251564,
      "grad_norm": 0.46484375,
      "learning_rate": 1.7257364896597842e-05,
      "loss": 0.8228,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2334
    },
    {
      "epoch": 0.3175142779439761,
      "grad_norm": 0.546875,
      "learning_rate": 1.7254097749263735e-05,
      "loss": 0.7017,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2335
    },
    {
      "epoch": 0.31765025836279576,
      "grad_norm": 0.4375,
      "learning_rate": 1.7250828966767303e-05,
      "loss": 0.5753,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2336
    },
    {
      "epoch": 0.31778623878161544,
      "grad_norm": 0.55078125,
      "learning_rate": 1.7247558549845372e-05,
      "loss": 0.5737,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2337
    },
    {
      "epoch": 0.3179222192004351,
      "grad_norm": 0.37890625,
      "learning_rate": 1.7244286499235136e-05,
      "loss": 0.7391,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2338
    },
    {
      "epoch": 0.3180581996192548,
      "grad_norm": 0.671875,
      "learning_rate": 1.724101281567415e-05,
      "loss": 0.6374,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2339
    },
    {
      "epoch": 0.31819418003807454,
      "grad_norm": 0.37109375,
      "learning_rate": 1.723773749990035e-05,
      "loss": 0.7876,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2340
    },
    {
      "epoch": 0.3183301604568942,
      "grad_norm": 0.302734375,
      "learning_rate": 1.7234460552652026e-05,
      "loss": 0.6221,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2341
    },
    {
      "epoch": 0.3184661408757139,
      "grad_norm": 0.294921875,
      "learning_rate": 1.7231181974667846e-05,
      "loss": 0.5804,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2342
    },
    {
      "epoch": 0.3186021212945336,
      "grad_norm": 0.44921875,
      "learning_rate": 1.7227901766686842e-05,
      "loss": 0.6864,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2343
    },
    {
      "epoch": 0.31873810171335326,
      "grad_norm": 0.349609375,
      "learning_rate": 1.7224619929448406e-05,
      "loss": 0.5566,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2344
    },
    {
      "epoch": 0.31887408213217294,
      "grad_norm": 0.31640625,
      "learning_rate": 1.722133646369231e-05,
      "loss": 0.431,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2345
    },
    {
      "epoch": 0.3190100625509927,
      "grad_norm": 0.34375,
      "learning_rate": 1.721805137015869e-05,
      "loss": 0.5804,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2346
    },
    {
      "epoch": 0.31914604296981236,
      "grad_norm": 0.87109375,
      "learning_rate": 1.721476464958804e-05,
      "loss": 0.6787,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2347
    },
    {
      "epoch": 0.31928202338863204,
      "grad_norm": 0.451171875,
      "learning_rate": 1.721147630272123e-05,
      "loss": 0.9093,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2348
    },
    {
      "epoch": 0.3194180038074517,
      "grad_norm": 0.400390625,
      "learning_rate": 1.7208186330299496e-05,
      "loss": 0.6087,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2349
    },
    {
      "epoch": 0.3195539842262714,
      "grad_norm": 0.25390625,
      "learning_rate": 1.7204894733064433e-05,
      "loss": 0.4141,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2350
    },
    {
      "epoch": 0.3196899646450911,
      "grad_norm": 0.69921875,
      "learning_rate": 1.720160151175801e-05,
      "loss": 0.7166,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2351
    },
    {
      "epoch": 0.3198259450639108,
      "grad_norm": 0.3203125,
      "learning_rate": 1.7198306667122557e-05,
      "loss": 0.5381,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2352
    },
    {
      "epoch": 0.3199619254827305,
      "grad_norm": 0.353515625,
      "learning_rate": 1.7195010199900773e-05,
      "loss": 0.5728,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2353
    },
    {
      "epoch": 0.3200979059015502,
      "grad_norm": 0.578125,
      "learning_rate": 1.7191712110835728e-05,
      "loss": 0.6512,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2354
    },
    {
      "epoch": 0.32023388632036986,
      "grad_norm": 0.291015625,
      "learning_rate": 1.7188412400670846e-05,
      "loss": 0.543,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2355
    },
    {
      "epoch": 0.32036986673918955,
      "grad_norm": 0.357421875,
      "learning_rate": 1.718511107014992e-05,
      "loss": 0.6255,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2356
    },
    {
      "epoch": 0.3205058471580092,
      "grad_norm": 0.953125,
      "learning_rate": 1.7181808120017115e-05,
      "loss": 0.5781,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2357
    },
    {
      "epoch": 0.32064182757682896,
      "grad_norm": 0.32421875,
      "learning_rate": 1.717850355101695e-05,
      "loss": 0.5679,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2358
    },
    {
      "epoch": 0.32077780799564865,
      "grad_norm": 0.3125,
      "learning_rate": 1.7175197363894326e-05,
      "loss": 0.5755,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2359
    },
    {
      "epoch": 0.3209137884144683,
      "grad_norm": 0.416015625,
      "learning_rate": 1.7171889559394488e-05,
      "loss": 0.6462,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2360
    },
    {
      "epoch": 0.321049768833288,
      "grad_norm": 0.458984375,
      "learning_rate": 1.7168580138263064e-05,
      "loss": 0.8636,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2361
    },
    {
      "epoch": 0.3211857492521077,
      "grad_norm": 0.390625,
      "learning_rate": 1.7165269101246026e-05,
      "loss": 0.7454,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2362
    },
    {
      "epoch": 0.32132172967092737,
      "grad_norm": 0.8984375,
      "learning_rate": 1.7161956449089736e-05,
      "loss": 0.7472,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2363
    },
    {
      "epoch": 0.32145771008974705,
      "grad_norm": 0.287109375,
      "learning_rate": 1.7158642182540896e-05,
      "loss": 0.4388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2364
    },
    {
      "epoch": 0.3215936905085668,
      "grad_norm": 0.37109375,
      "learning_rate": 1.7155326302346584e-05,
      "loss": 0.6234,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2365
    },
    {
      "epoch": 0.32172967092738647,
      "grad_norm": 0.359375,
      "learning_rate": 1.7152008809254247e-05,
      "loss": 0.6182,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2366
    },
    {
      "epoch": 0.32186565134620615,
      "grad_norm": 0.443359375,
      "learning_rate": 1.7148689704011682e-05,
      "loss": 0.8507,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2367
    },
    {
      "epoch": 0.32200163176502583,
      "grad_norm": 0.24609375,
      "learning_rate": 1.714536898736705e-05,
      "loss": 0.4012,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2368
    },
    {
      "epoch": 0.3221376121838455,
      "grad_norm": 0.416015625,
      "learning_rate": 1.7142046660068898e-05,
      "loss": 0.5487,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2369
    },
    {
      "epoch": 0.3222735926026652,
      "grad_norm": 0.38671875,
      "learning_rate": 1.7138722722866106e-05,
      "loss": 0.7354,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2370
    },
    {
      "epoch": 0.32240957302148493,
      "grad_norm": 0.55859375,
      "learning_rate": 1.7135397176507935e-05,
      "loss": 0.563,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2371
    },
    {
      "epoch": 0.3225455534403046,
      "grad_norm": 0.96875,
      "learning_rate": 1.7132070021744e-05,
      "loss": 0.8333,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2372
    },
    {
      "epoch": 0.3226815338591243,
      "grad_norm": 0.298828125,
      "learning_rate": 1.7128741259324283e-05,
      "loss": 0.4261,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2373
    },
    {
      "epoch": 0.322817514277944,
      "grad_norm": 0.427734375,
      "learning_rate": 1.7125410889999135e-05,
      "loss": 0.7869,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2374
    },
    {
      "epoch": 0.32295349469676365,
      "grad_norm": 0.35546875,
      "learning_rate": 1.7122078914519257e-05,
      "loss": 0.5627,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2375
    },
    {
      "epoch": 0.32308947511558334,
      "grad_norm": 0.353515625,
      "learning_rate": 1.7118745333635713e-05,
      "loss": 0.6208,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2376
    },
    {
      "epoch": 0.32322545553440307,
      "grad_norm": 0.5234375,
      "learning_rate": 1.7115410148099933e-05,
      "loss": 0.7842,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2377
    },
    {
      "epoch": 0.32336143595322275,
      "grad_norm": 0.423828125,
      "learning_rate": 1.7112073358663716e-05,
      "loss": 0.8057,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2378
    },
    {
      "epoch": 0.32349741637204243,
      "grad_norm": 0.455078125,
      "learning_rate": 1.7108734966079214e-05,
      "loss": 0.6276,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2379
    },
    {
      "epoch": 0.3236333967908621,
      "grad_norm": 0.416015625,
      "learning_rate": 1.7105394971098938e-05,
      "loss": 0.6134,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2380
    },
    {
      "epoch": 0.3237693772096818,
      "grad_norm": 0.361328125,
      "learning_rate": 1.710205337447576e-05,
      "loss": 0.4193,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2381
    },
    {
      "epoch": 0.3239053576285015,
      "grad_norm": 0.6875,
      "learning_rate": 1.7098710176962923e-05,
      "loss": 0.4982,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2382
    },
    {
      "epoch": 0.32404133804732116,
      "grad_norm": 0.380859375,
      "learning_rate": 1.7095365379314025e-05,
      "loss": 0.617,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2383
    },
    {
      "epoch": 0.3241773184661409,
      "grad_norm": 0.228515625,
      "learning_rate": 1.709201898228302e-05,
      "loss": 0.4136,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2384
    },
    {
      "epoch": 0.3243132988849606,
      "grad_norm": 0.703125,
      "learning_rate": 1.708867098662423e-05,
      "loss": 0.6724,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2385
    },
    {
      "epoch": 0.32444927930378026,
      "grad_norm": 0.46484375,
      "learning_rate": 1.7085321393092332e-05,
      "loss": 0.8628,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2386
    },
    {
      "epoch": 0.32458525972259994,
      "grad_norm": 0.32421875,
      "learning_rate": 1.7081970202442363e-05,
      "loss": 0.5786,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2387
    },
    {
      "epoch": 0.3247212401414196,
      "grad_norm": 0.30859375,
      "learning_rate": 1.707861741542973e-05,
      "loss": 0.6346,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2388
    },
    {
      "epoch": 0.3248572205602393,
      "grad_norm": 0.60546875,
      "learning_rate": 1.7075263032810188e-05,
      "loss": 0.745,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2389
    },
    {
      "epoch": 0.32499320097905904,
      "grad_norm": 0.279296875,
      "learning_rate": 1.7071907055339855e-05,
      "loss": 0.4753,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2390
    },
    {
      "epoch": 0.3251291813978787,
      "grad_norm": 0.51953125,
      "learning_rate": 1.706854948377521e-05,
      "loss": 0.5228,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2391
    },
    {
      "epoch": 0.3252651618166984,
      "grad_norm": 0.271484375,
      "learning_rate": 1.706519031887309e-05,
      "loss": 0.5309,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2392
    },
    {
      "epoch": 0.3254011422355181,
      "grad_norm": 0.31640625,
      "learning_rate": 1.7061829561390692e-05,
      "loss": 0.6768,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2393
    },
    {
      "epoch": 0.32553712265433776,
      "grad_norm": 0.52734375,
      "learning_rate": 1.7058467212085574e-05,
      "loss": 0.8018,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2394
    },
    {
      "epoch": 0.32567310307315744,
      "grad_norm": 0.46875,
      "learning_rate": 1.7055103271715647e-05,
      "loss": 0.707,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2395
    },
    {
      "epoch": 0.3258090834919772,
      "grad_norm": 0.357421875,
      "learning_rate": 1.7051737741039187e-05,
      "loss": 0.6274,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2396
    },
    {
      "epoch": 0.32594506391079686,
      "grad_norm": 0.373046875,
      "learning_rate": 1.7048370620814827e-05,
      "loss": 0.6458,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2397
    },
    {
      "epoch": 0.32608104432961654,
      "grad_norm": 0.3203125,
      "learning_rate": 1.704500191180155e-05,
      "loss": 0.6076,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2398
    },
    {
      "epoch": 0.3262170247484362,
      "grad_norm": 0.353515625,
      "learning_rate": 1.7041631614758716e-05,
      "loss": 0.6414,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2399
    },
    {
      "epoch": 0.3263530051672559,
      "grad_norm": 0.328125,
      "learning_rate": 1.703825973044602e-05,
      "loss": 0.6787,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2400
    },
    {
      "epoch": 0.3264889855860756,
      "grad_norm": 0.60546875,
      "learning_rate": 1.7034886259623534e-05,
      "loss": 0.6296,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2401
    },
    {
      "epoch": 0.32662496600489527,
      "grad_norm": 0.3828125,
      "learning_rate": 1.7031511203051675e-05,
      "loss": 0.6807,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2402
    },
    {
      "epoch": 0.326760946423715,
      "grad_norm": 0.2353515625,
      "learning_rate": 1.7028134561491222e-05,
      "loss": 0.4357,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2403
    },
    {
      "epoch": 0.3268969268425347,
      "grad_norm": 0.3359375,
      "learning_rate": 1.702475633570331e-05,
      "loss": 0.6722,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2404
    },
    {
      "epoch": 0.32703290726135437,
      "grad_norm": 0.484375,
      "learning_rate": 1.702137652644944e-05,
      "loss": 0.7891,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2405
    },
    {
      "epoch": 0.32716888768017405,
      "grad_norm": 0.53515625,
      "learning_rate": 1.701799513449145e-05,
      "loss": 0.8117,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2406
    },
    {
      "epoch": 0.32730486809899373,
      "grad_norm": 0.326171875,
      "learning_rate": 1.7014612160591556e-05,
      "loss": 0.5703,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2407
    },
    {
      "epoch": 0.3274408485178134,
      "grad_norm": 0.404296875,
      "learning_rate": 1.7011227605512325e-05,
      "loss": 0.7074,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2408
    },
    {
      "epoch": 0.32757682893663315,
      "grad_norm": 0.625,
      "learning_rate": 1.7007841470016667e-05,
      "loss": 0.6654,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2409
    },
    {
      "epoch": 0.3277128093554528,
      "grad_norm": 0.54296875,
      "learning_rate": 1.7004453754867866e-05,
      "loss": 0.3226,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2410
    },
    {
      "epoch": 0.3278487897742725,
      "grad_norm": 0.357421875,
      "learning_rate": 1.700106446082955e-05,
      "loss": 0.5833,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2411
    },
    {
      "epoch": 0.3279847701930922,
      "grad_norm": 2.328125,
      "learning_rate": 1.699767358866571e-05,
      "loss": 0.8271,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2412
    },
    {
      "epoch": 0.32812075061191187,
      "grad_norm": 0.8828125,
      "learning_rate": 1.699428113914069e-05,
      "loss": 0.8468,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2413
    },
    {
      "epoch": 0.32825673103073155,
      "grad_norm": 0.71875,
      "learning_rate": 1.6990887113019184e-05,
      "loss": 0.7345,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2414
    },
    {
      "epoch": 0.3283927114495513,
      "grad_norm": 0.427734375,
      "learning_rate": 1.6987491511066258e-05,
      "loss": 0.6855,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2415
    },
    {
      "epoch": 0.32852869186837097,
      "grad_norm": 0.2734375,
      "learning_rate": 1.6984094334047313e-05,
      "loss": 0.4523,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2416
    },
    {
      "epoch": 0.32866467228719065,
      "grad_norm": 0.275390625,
      "learning_rate": 1.6980695582728115e-05,
      "loss": 0.5337,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2417
    },
    {
      "epoch": 0.32880065270601033,
      "grad_norm": 0.333984375,
      "learning_rate": 1.6977295257874787e-05,
      "loss": 0.5483,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2418
    },
    {
      "epoch": 0.32893663312483,
      "grad_norm": 0.3203125,
      "learning_rate": 1.6973893360253807e-05,
      "loss": 0.6646,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2419
    },
    {
      "epoch": 0.3290726135436497,
      "grad_norm": 0.375,
      "learning_rate": 1.6970489890631996e-05,
      "loss": 0.6457,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2420
    },
    {
      "epoch": 0.32920859396246943,
      "grad_norm": 0.55859375,
      "learning_rate": 1.6967084849776543e-05,
      "loss": 0.7974,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2421
    },
    {
      "epoch": 0.3293445743812891,
      "grad_norm": 0.30859375,
      "learning_rate": 1.6963678238454982e-05,
      "loss": 0.6608,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2422
    },
    {
      "epoch": 0.3294805548001088,
      "grad_norm": 0.7890625,
      "learning_rate": 1.6960270057435212e-05,
      "loss": 0.6769,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2423
    },
    {
      "epoch": 0.3296165352189285,
      "grad_norm": 0.373046875,
      "learning_rate": 1.6956860307485473e-05,
      "loss": 0.6502,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2424
    },
    {
      "epoch": 0.32975251563774816,
      "grad_norm": 0.4375,
      "learning_rate": 1.6953448989374365e-05,
      "loss": 0.5676,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2425
    },
    {
      "epoch": 0.32988849605656784,
      "grad_norm": 0.291015625,
      "learning_rate": 1.695003610387084e-05,
      "loss": 0.4684,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2426
    },
    {
      "epoch": 0.3300244764753875,
      "grad_norm": 0.400390625,
      "learning_rate": 1.6946621651744208e-05,
      "loss": 0.8175,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2427
    },
    {
      "epoch": 0.33016045689420725,
      "grad_norm": 0.40625,
      "learning_rate": 1.6943205633764123e-05,
      "loss": 0.7331,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2428
    },
    {
      "epoch": 0.33029643731302694,
      "grad_norm": 0.9296875,
      "learning_rate": 1.69397880507006e-05,
      "loss": 0.7697,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2429
    },
    {
      "epoch": 0.3304324177318466,
      "grad_norm": 0.51953125,
      "learning_rate": 1.6936368903324004e-05,
      "loss": 0.6497,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2430
    },
    {
      "epoch": 0.3305683981506663,
      "grad_norm": 0.33984375,
      "learning_rate": 1.6932948192405055e-05,
      "loss": 0.578,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2431
    },
    {
      "epoch": 0.330704378569486,
      "grad_norm": 0.6953125,
      "learning_rate": 1.692952591871482e-05,
      "loss": 0.8696,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2432
    },
    {
      "epoch": 0.33084035898830566,
      "grad_norm": 0.451171875,
      "learning_rate": 1.6926102083024722e-05,
      "loss": 0.4784,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2433
    },
    {
      "epoch": 0.3309763394071254,
      "grad_norm": 0.486328125,
      "learning_rate": 1.6922676686106537e-05,
      "loss": 0.7728,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2434
    },
    {
      "epoch": 0.3311123198259451,
      "grad_norm": 0.28125,
      "learning_rate": 1.691924972873239e-05,
      "loss": 0.5137,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2435
    },
    {
      "epoch": 0.33124830024476476,
      "grad_norm": 0.365234375,
      "learning_rate": 1.6915821211674752e-05,
      "loss": 0.474,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2436
    },
    {
      "epoch": 0.33138428066358444,
      "grad_norm": 0.796875,
      "learning_rate": 1.691239113570646e-05,
      "loss": 0.7609,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2437
    },
    {
      "epoch": 0.3315202610824041,
      "grad_norm": 0.359375,
      "learning_rate": 1.6908959501600702e-05,
      "loss": 0.5923,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2438
    },
    {
      "epoch": 0.3316562415012238,
      "grad_norm": 0.41015625,
      "learning_rate": 1.6905526310131e-05,
      "loss": 0.702,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2439
    },
    {
      "epoch": 0.33179222192004354,
      "grad_norm": 0.5859375,
      "learning_rate": 1.690209156207124e-05,
      "loss": 0.8556,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2440
    },
    {
      "epoch": 0.3319282023388632,
      "grad_norm": 0.56640625,
      "learning_rate": 1.6898655258195654e-05,
      "loss": 0.7351,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2441
    },
    {
      "epoch": 0.3320641827576829,
      "grad_norm": 0.66015625,
      "learning_rate": 1.6895217399278833e-05,
      "loss": 0.8136,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2442
    },
    {
      "epoch": 0.3322001631765026,
      "grad_norm": 0.228515625,
      "learning_rate": 1.6891777986095706e-05,
      "loss": 0.3377,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2443
    },
    {
      "epoch": 0.33233614359532226,
      "grad_norm": 0.390625,
      "learning_rate": 1.6888337019421564e-05,
      "loss": 0.682,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2444
    },
    {
      "epoch": 0.33247212401414195,
      "grad_norm": 0.392578125,
      "learning_rate": 1.6884894500032038e-05,
      "loss": 0.5682,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2445
    },
    {
      "epoch": 0.3326081044329616,
      "grad_norm": 0.50390625,
      "learning_rate": 1.6881450428703118e-05,
      "loss": 0.8359,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2446
    },
    {
      "epoch": 0.33274408485178136,
      "grad_norm": 0.67578125,
      "learning_rate": 1.687800480621114e-05,
      "loss": 0.7502,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2447
    },
    {
      "epoch": 0.33288006527060104,
      "grad_norm": 0.81640625,
      "learning_rate": 1.6874557633332784e-05,
      "loss": 0.6213,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2448
    },
    {
      "epoch": 0.3330160456894207,
      "grad_norm": 0.65234375,
      "learning_rate": 1.687110891084509e-05,
      "loss": 0.7944,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2449
    },
    {
      "epoch": 0.3331520261082404,
      "grad_norm": 0.494140625,
      "learning_rate": 1.6867658639525443e-05,
      "loss": 0.6833,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2450
    },
    {
      "epoch": 0.3332880065270601,
      "grad_norm": 0.384765625,
      "learning_rate": 1.686420682015157e-05,
      "loss": 0.7544,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2451
    },
    {
      "epoch": 0.33342398694587977,
      "grad_norm": 0.333984375,
      "learning_rate": 1.686075345350156e-05,
      "loss": 0.681,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2452
    },
    {
      "epoch": 0.3335599673646995,
      "grad_norm": 0.65234375,
      "learning_rate": 1.685729854035384e-05,
      "loss": 0.6812,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2453
    },
    {
      "epoch": 0.3336959477835192,
      "grad_norm": 0.55859375,
      "learning_rate": 1.685384208148719e-05,
      "loss": 0.5968,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2454
    },
    {
      "epoch": 0.33383192820233887,
      "grad_norm": 0.30078125,
      "learning_rate": 1.685038407768074e-05,
      "loss": 0.5863,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2455
    },
    {
      "epoch": 0.33396790862115855,
      "grad_norm": 0.6953125,
      "learning_rate": 1.684692452971397e-05,
      "loss": 0.7419,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2456
    },
    {
      "epoch": 0.33410388903997823,
      "grad_norm": 0.21484375,
      "learning_rate": 1.6843463438366697e-05,
      "loss": 0.3351,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2457
    },
    {
      "epoch": 0.3342398694587979,
      "grad_norm": 0.30859375,
      "learning_rate": 1.6840000804419094e-05,
      "loss": 0.4541,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2458
    },
    {
      "epoch": 0.33437584987761765,
      "grad_norm": 0.345703125,
      "learning_rate": 1.6836536628651688e-05,
      "loss": 0.6069,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2459
    },
    {
      "epoch": 0.33451183029643733,
      "grad_norm": 0.376953125,
      "learning_rate": 1.6833070911845335e-05,
      "loss": 0.7435,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2460
    },
    {
      "epoch": 0.334647810715257,
      "grad_norm": 0.51171875,
      "learning_rate": 1.682960365478126e-05,
      "loss": 0.6746,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2461
    },
    {
      "epoch": 0.3347837911340767,
      "grad_norm": 0.484375,
      "learning_rate": 1.6826134858241027e-05,
      "loss": 0.771,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2462
    },
    {
      "epoch": 0.3349197715528964,
      "grad_norm": 0.62109375,
      "learning_rate": 1.6822664523006533e-05,
      "loss": 0.5516,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2463
    },
    {
      "epoch": 0.33505575197171605,
      "grad_norm": 0.6640625,
      "learning_rate": 1.6819192649860047e-05,
      "loss": 0.507,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2464
    },
    {
      "epoch": 0.3351917323905358,
      "grad_norm": 0.333984375,
      "learning_rate": 1.681571923958416e-05,
      "loss": 0.5713,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2465
    },
    {
      "epoch": 0.33532771280935547,
      "grad_norm": 0.70703125,
      "learning_rate": 1.6812244292961827e-05,
      "loss": 0.7118,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2466
    },
    {
      "epoch": 0.33546369322817515,
      "grad_norm": 0.515625,
      "learning_rate": 1.6808767810776346e-05,
      "loss": 0.7889,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2467
    },
    {
      "epoch": 0.33559967364699483,
      "grad_norm": 0.435546875,
      "learning_rate": 1.6805289793811356e-05,
      "loss": 0.6097,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2468
    },
    {
      "epoch": 0.3357356540658145,
      "grad_norm": 0.333984375,
      "learning_rate": 1.6801810242850844e-05,
      "loss": 0.5775,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2469
    },
    {
      "epoch": 0.3358716344846342,
      "grad_norm": 0.31640625,
      "learning_rate": 1.679832915867914e-05,
      "loss": 0.4359,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2470
    },
    {
      "epoch": 0.3360076149034539,
      "grad_norm": 0.486328125,
      "learning_rate": 1.6794846542080937e-05,
      "loss": 0.6702,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2471
    },
    {
      "epoch": 0.3361435953222736,
      "grad_norm": 0.376953125,
      "learning_rate": 1.6791362393841243e-05,
      "loss": 0.5252,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2472
    },
    {
      "epoch": 0.3362795757410933,
      "grad_norm": 0.87109375,
      "learning_rate": 1.6787876714745432e-05,
      "loss": 0.7181,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2473
    },
    {
      "epoch": 0.336415556159913,
      "grad_norm": 0.69921875,
      "learning_rate": 1.6784389505579222e-05,
      "loss": 0.791,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2474
    },
    {
      "epoch": 0.33655153657873266,
      "grad_norm": 0.62890625,
      "learning_rate": 1.678090076712867e-05,
      "loss": 0.6232,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2475
    },
    {
      "epoch": 0.33668751699755234,
      "grad_norm": 0.61328125,
      "learning_rate": 1.6777410500180183e-05,
      "loss": 0.7103,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2476
    },
    {
      "epoch": 0.336823497416372,
      "grad_norm": 0.5703125,
      "learning_rate": 1.6773918705520506e-05,
      "loss": 0.889,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2477
    },
    {
      "epoch": 0.33695947783519176,
      "grad_norm": 0.3359375,
      "learning_rate": 1.6770425383936734e-05,
      "loss": 0.5843,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2478
    },
    {
      "epoch": 0.33709545825401144,
      "grad_norm": 0.462890625,
      "learning_rate": 1.6766930536216307e-05,
      "loss": 0.7062,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2479
    },
    {
      "epoch": 0.3372314386728311,
      "grad_norm": 0.267578125,
      "learning_rate": 1.6763434163147e-05,
      "loss": 0.5531,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2480
    },
    {
      "epoch": 0.3373674190916508,
      "grad_norm": 0.353515625,
      "learning_rate": 1.675993626551695e-05,
      "loss": 0.627,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2481
    },
    {
      "epoch": 0.3375033995104705,
      "grad_norm": 3.71875,
      "learning_rate": 1.6756436844114613e-05,
      "loss": 0.7417,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2482
    },
    {
      "epoch": 0.33763937992929016,
      "grad_norm": 0.69140625,
      "learning_rate": 1.6752935899728808e-05,
      "loss": 0.7539,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2483
    },
    {
      "epoch": 0.3377753603481099,
      "grad_norm": 0.55859375,
      "learning_rate": 1.674943343314869e-05,
      "loss": 0.5483,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2484
    },
    {
      "epoch": 0.3379113407669296,
      "grad_norm": 0.28125,
      "learning_rate": 1.6745929445163757e-05,
      "loss": 0.6024,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2485
    },
    {
      "epoch": 0.33804732118574926,
      "grad_norm": 0.87109375,
      "learning_rate": 1.674242393656385e-05,
      "loss": 0.6911,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2486
    },
    {
      "epoch": 0.33818330160456894,
      "grad_norm": 0.486328125,
      "learning_rate": 1.673891690813915e-05,
      "loss": 0.6799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2487
    },
    {
      "epoch": 0.3383192820233886,
      "grad_norm": 0.59375,
      "learning_rate": 1.6735408360680193e-05,
      "loss": 0.7487,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2488
    },
    {
      "epoch": 0.3384552624422083,
      "grad_norm": 0.58203125,
      "learning_rate": 1.673189829497784e-05,
      "loss": 0.6987,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2489
    },
    {
      "epoch": 0.338591242861028,
      "grad_norm": 0.74609375,
      "learning_rate": 1.6728386711823313e-05,
      "loss": 0.6479,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2490
    },
    {
      "epoch": 0.3387272232798477,
      "grad_norm": 0.345703125,
      "learning_rate": 1.6724873612008155e-05,
      "loss": 0.6247,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2491
    },
    {
      "epoch": 0.3388632036986674,
      "grad_norm": 0.45703125,
      "learning_rate": 1.672135899632427e-05,
      "loss": 0.7629,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2492
    },
    {
      "epoch": 0.3389991841174871,
      "grad_norm": 0.34765625,
      "learning_rate": 1.6717842865563886e-05,
      "loss": 0.6278,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2493
    },
    {
      "epoch": 0.33913516453630677,
      "grad_norm": 0.3671875,
      "learning_rate": 1.671432522051959e-05,
      "loss": 0.6621,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2494
    },
    {
      "epoch": 0.33927114495512645,
      "grad_norm": 0.412109375,
      "learning_rate": 1.6710806061984298e-05,
      "loss": 0.7808,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2495
    },
    {
      "epoch": 0.33940712537394613,
      "grad_norm": 0.328125,
      "learning_rate": 1.6707285390751275e-05,
      "loss": 0.575,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2496
    },
    {
      "epoch": 0.33954310579276586,
      "grad_norm": 0.5234375,
      "learning_rate": 1.6703763207614118e-05,
      "loss": 0.7884,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2497
    },
    {
      "epoch": 0.33967908621158555,
      "grad_norm": 0.419921875,
      "learning_rate": 1.6700239513366777e-05,
      "loss": 0.7782,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2498
    },
    {
      "epoch": 0.3398150666304052,
      "grad_norm": 0.376953125,
      "learning_rate": 1.669671430880353e-05,
      "loss": 0.7259,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2499
    },
    {
      "epoch": 0.3399510470492249,
      "grad_norm": 0.51953125,
      "learning_rate": 1.6693187594719002e-05,
      "loss": 0.5929,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2500
    },
    {
      "epoch": 0.3400870274680446,
      "grad_norm": 0.50390625,
      "learning_rate": 1.668965937190816e-05,
      "loss": 0.8244,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2501
    },
    {
      "epoch": 0.34022300788686427,
      "grad_norm": 0.53515625,
      "learning_rate": 1.6686129641166308e-05,
      "loss": 0.5314,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2502
    },
    {
      "epoch": 0.340358988305684,
      "grad_norm": 0.375,
      "learning_rate": 1.6682598403289086e-05,
      "loss": 0.5612,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2503
    },
    {
      "epoch": 0.3404949687245037,
      "grad_norm": 0.306640625,
      "learning_rate": 1.6679065659072486e-05,
      "loss": 0.5797,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2504
    },
    {
      "epoch": 0.34063094914332337,
      "grad_norm": 0.373046875,
      "learning_rate": 1.6675531409312827e-05,
      "loss": 0.6746,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2505
    },
    {
      "epoch": 0.34076692956214305,
      "grad_norm": 0.326171875,
      "learning_rate": 1.6671995654806774e-05,
      "loss": 0.6318,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2506
    },
    {
      "epoch": 0.34090290998096273,
      "grad_norm": 0.51953125,
      "learning_rate": 1.6668458396351327e-05,
      "loss": 0.8089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2507
    },
    {
      "epoch": 0.3410388903997824,
      "grad_norm": 0.3828125,
      "learning_rate": 1.6664919634743826e-05,
      "loss": 0.5339,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2508
    },
    {
      "epoch": 0.3411748708186021,
      "grad_norm": 0.828125,
      "learning_rate": 1.6661379370781956e-05,
      "loss": 0.7388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2509
    },
    {
      "epoch": 0.34131085123742183,
      "grad_norm": 0.267578125,
      "learning_rate": 1.6657837605263733e-05,
      "loss": 0.5186,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2510
    },
    {
      "epoch": 0.3414468316562415,
      "grad_norm": 0.34765625,
      "learning_rate": 1.6654294338987513e-05,
      "loss": 0.8431,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2511
    },
    {
      "epoch": 0.3415828120750612,
      "grad_norm": 0.33203125,
      "learning_rate": 1.6650749572751994e-05,
      "loss": 0.5931,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2512
    },
    {
      "epoch": 0.3417187924938809,
      "grad_norm": 0.578125,
      "learning_rate": 1.664720330735621e-05,
      "loss": 0.7723,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2513
    },
    {
      "epoch": 0.34185477291270056,
      "grad_norm": 0.30078125,
      "learning_rate": 1.664365554359953e-05,
      "loss": 0.7217,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2514
    },
    {
      "epoch": 0.34199075333152024,
      "grad_norm": 0.37109375,
      "learning_rate": 1.6640106282281665e-05,
      "loss": 0.5986,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2515
    },
    {
      "epoch": 0.34212673375034,
      "grad_norm": 0.59375,
      "learning_rate": 1.663655552420266e-05,
      "loss": 0.7482,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2516
    },
    {
      "epoch": 0.34226271416915965,
      "grad_norm": 0.640625,
      "learning_rate": 1.6633003270162903e-05,
      "loss": 0.6934,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2517
    },
    {
      "epoch": 0.34239869458797934,
      "grad_norm": 0.48828125,
      "learning_rate": 1.6629449520963112e-05,
      "loss": 0.3488,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2518
    },
    {
      "epoch": 0.342534675006799,
      "grad_norm": 0.29296875,
      "learning_rate": 1.6625894277404347e-05,
      "loss": 0.5462,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2519
    },
    {
      "epoch": 0.3426706554256187,
      "grad_norm": 0.40625,
      "learning_rate": 1.6622337540288005e-05,
      "loss": 0.6195,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2520
    },
    {
      "epoch": 0.3428066358444384,
      "grad_norm": 0.443359375,
      "learning_rate": 1.6618779310415813e-05,
      "loss": 0.5688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2521
    },
    {
      "epoch": 0.3429426162632581,
      "grad_norm": 0.54296875,
      "learning_rate": 1.6615219588589844e-05,
      "loss": 0.7153,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2522
    },
    {
      "epoch": 0.3430785966820778,
      "grad_norm": 0.51953125,
      "learning_rate": 1.6611658375612502e-05,
      "loss": 0.8218,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2523
    },
    {
      "epoch": 0.3432145771008975,
      "grad_norm": 0.326171875,
      "learning_rate": 1.6608095672286527e-05,
      "loss": 0.529,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2524
    },
    {
      "epoch": 0.34335055751971716,
      "grad_norm": 0.419921875,
      "learning_rate": 1.6604531479414998e-05,
      "loss": 0.7956,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2525
    },
    {
      "epoch": 0.34348653793853684,
      "grad_norm": 0.455078125,
      "learning_rate": 1.6600965797801322e-05,
      "loss": 0.8755,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2526
    },
    {
      "epoch": 0.3436225183573565,
      "grad_norm": 0.359375,
      "learning_rate": 1.6597398628249256e-05,
      "loss": 0.5993,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2527
    },
    {
      "epoch": 0.34375849877617626,
      "grad_norm": 0.32421875,
      "learning_rate": 1.6593829971562877e-05,
      "loss": 0.6162,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2528
    },
    {
      "epoch": 0.34389447919499594,
      "grad_norm": 0.5703125,
      "learning_rate": 1.6590259828546603e-05,
      "loss": 0.7383,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2529
    },
    {
      "epoch": 0.3440304596138156,
      "grad_norm": 0.51171875,
      "learning_rate": 1.6586688200005193e-05,
      "loss": 0.5588,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2530
    },
    {
      "epoch": 0.3441664400326353,
      "grad_norm": 0.625,
      "learning_rate": 1.6583115086743734e-05,
      "loss": 0.6523,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2531
    },
    {
      "epoch": 0.344302420451455,
      "grad_norm": 0.458984375,
      "learning_rate": 1.6579540489567648e-05,
      "loss": 0.772,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2532
    },
    {
      "epoch": 0.34443840087027466,
      "grad_norm": 0.71484375,
      "learning_rate": 1.6575964409282695e-05,
      "loss": 0.5356,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2533
    },
    {
      "epoch": 0.34457438128909434,
      "grad_norm": 0.49609375,
      "learning_rate": 1.6572386846694963e-05,
      "loss": 0.8172,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2534
    },
    {
      "epoch": 0.3447103617079141,
      "grad_norm": 0.349609375,
      "learning_rate": 1.6568807802610886e-05,
      "loss": 0.5802,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2535
    },
    {
      "epoch": 0.34484634212673376,
      "grad_norm": 0.462890625,
      "learning_rate": 1.6565227277837214e-05,
      "loss": 0.742,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2536
    },
    {
      "epoch": 0.34498232254555344,
      "grad_norm": 0.265625,
      "learning_rate": 1.656164527318105e-05,
      "loss": 0.4155,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2537
    },
    {
      "epoch": 0.3451183029643731,
      "grad_norm": 2.21875,
      "learning_rate": 1.6558061789449818e-05,
      "loss": 1.007,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2538
    },
    {
      "epoch": 0.3452542833831928,
      "grad_norm": 0.482421875,
      "learning_rate": 1.6554476827451277e-05,
      "loss": 0.7458,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2539
    },
    {
      "epoch": 0.3453902638020125,
      "grad_norm": 0.4375,
      "learning_rate": 1.6550890387993525e-05,
      "loss": 0.8304,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2540
    },
    {
      "epoch": 0.3455262442208322,
      "grad_norm": 0.5234375,
      "learning_rate": 1.654730247188499e-05,
      "loss": 0.5679,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2541
    },
    {
      "epoch": 0.3456622246396519,
      "grad_norm": 0.61328125,
      "learning_rate": 1.6543713079934426e-05,
      "loss": 0.7515,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2542
    },
    {
      "epoch": 0.3457982050584716,
      "grad_norm": 0.419921875,
      "learning_rate": 1.654012221295093e-05,
      "loss": 0.501,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2543
    },
    {
      "epoch": 0.34593418547729127,
      "grad_norm": 0.333984375,
      "learning_rate": 1.653652987174393e-05,
      "loss": 0.6842,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2544
    },
    {
      "epoch": 0.34607016589611095,
      "grad_norm": 0.57421875,
      "learning_rate": 1.653293605712318e-05,
      "loss": 0.8055,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2545
    },
    {
      "epoch": 0.34620614631493063,
      "grad_norm": 0.380859375,
      "learning_rate": 1.652934076989877e-05,
      "loss": 0.5649,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2546
    },
    {
      "epoch": 0.34634212673375037,
      "grad_norm": 0.3125,
      "learning_rate": 1.6525744010881125e-05,
      "loss": 0.5708,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2547
    },
    {
      "epoch": 0.34647810715257005,
      "grad_norm": 0.359375,
      "learning_rate": 1.652214578088099e-05,
      "loss": 0.4259,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2548
    },
    {
      "epoch": 0.34661408757138973,
      "grad_norm": 0.443359375,
      "learning_rate": 1.651854608070946e-05,
      "loss": 0.6742,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2549
    },
    {
      "epoch": 0.3467500679902094,
      "grad_norm": 0.34765625,
      "learning_rate": 1.6514944911177946e-05,
      "loss": 0.5848,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2550
    },
    {
      "epoch": 0.3468860484090291,
      "grad_norm": 0.419921875,
      "learning_rate": 1.6511342273098195e-05,
      "loss": 0.679,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2551
    },
    {
      "epoch": 0.34702202882784877,
      "grad_norm": 0.373046875,
      "learning_rate": 1.6507738167282294e-05,
      "loss": 0.6113,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2552
    },
    {
      "epoch": 0.34715800924666845,
      "grad_norm": 0.5078125,
      "learning_rate": 1.650413259454264e-05,
      "loss": 0.6143,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2553
    },
    {
      "epoch": 0.3472939896654882,
      "grad_norm": 0.498046875,
      "learning_rate": 1.650052555569198e-05,
      "loss": 0.701,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2554
    },
    {
      "epoch": 0.34742997008430787,
      "grad_norm": 0.6171875,
      "learning_rate": 1.6496917051543386e-05,
      "loss": 0.7537,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2555
    },
    {
      "epoch": 0.34756595050312755,
      "grad_norm": 0.40625,
      "learning_rate": 1.649330708291025e-05,
      "loss": 0.6644,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2556
    },
    {
      "epoch": 0.34770193092194723,
      "grad_norm": 0.310546875,
      "learning_rate": 1.6489695650606314e-05,
      "loss": 0.637,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2557
    },
    {
      "epoch": 0.3478379113407669,
      "grad_norm": 0.369140625,
      "learning_rate": 1.6486082755445635e-05,
      "loss": 0.5994,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2558
    },
    {
      "epoch": 0.3479738917595866,
      "grad_norm": 0.412109375,
      "learning_rate": 1.6482468398242606e-05,
      "loss": 0.5921,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2559
    },
    {
      "epoch": 0.34810987217840633,
      "grad_norm": 0.50390625,
      "learning_rate": 1.6478852579811943e-05,
      "loss": 0.5596,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2560
    },
    {
      "epoch": 0.348245852597226,
      "grad_norm": 0.419921875,
      "learning_rate": 1.6475235300968696e-05,
      "loss": 0.6891,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2561
    },
    {
      "epoch": 0.3483818330160457,
      "grad_norm": 0.2578125,
      "learning_rate": 1.647161656252825e-05,
      "loss": 0.4948,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2562
    },
    {
      "epoch": 0.3485178134348654,
      "grad_norm": 0.328125,
      "learning_rate": 1.6467996365306304e-05,
      "loss": 0.5894,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2563
    },
    {
      "epoch": 0.34865379385368506,
      "grad_norm": 0.283203125,
      "learning_rate": 1.64643747101189e-05,
      "loss": 0.5549,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2564
    },
    {
      "epoch": 0.34878977427250474,
      "grad_norm": 0.74609375,
      "learning_rate": 1.6460751597782402e-05,
      "loss": 0.7253,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2565
    },
    {
      "epoch": 0.3489257546913245,
      "grad_norm": 0.40625,
      "learning_rate": 1.6457127029113508e-05,
      "loss": 0.8138,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2566
    },
    {
      "epoch": 0.34906173511014416,
      "grad_norm": 0.3125,
      "learning_rate": 1.645350100492923e-05,
      "loss": 0.6455,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2567
    },
    {
      "epoch": 0.34919771552896384,
      "grad_norm": 0.73828125,
      "learning_rate": 1.6449873526046928e-05,
      "loss": 0.8564,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2568
    },
    {
      "epoch": 0.3493336959477835,
      "grad_norm": 0.39453125,
      "learning_rate": 1.6446244593284277e-05,
      "loss": 0.5665,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2569
    },
    {
      "epoch": 0.3494696763666032,
      "grad_norm": 0.373046875,
      "learning_rate": 1.644261420745928e-05,
      "loss": 0.7184,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2570
    },
    {
      "epoch": 0.3496056567854229,
      "grad_norm": 0.5078125,
      "learning_rate": 1.6438982369390273e-05,
      "loss": 0.6554,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2571
    },
    {
      "epoch": 0.34974163720424256,
      "grad_norm": 0.5078125,
      "learning_rate": 1.6435349079895918e-05,
      "loss": 0.8664,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2572
    },
    {
      "epoch": 0.3498776176230623,
      "grad_norm": 0.46875,
      "learning_rate": 1.64317143397952e-05,
      "loss": 0.5039,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2573
    },
    {
      "epoch": 0.350013598041882,
      "grad_norm": 0.326171875,
      "learning_rate": 1.642807814990744e-05,
      "loss": 0.5815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2574
    },
    {
      "epoch": 0.35014957846070166,
      "grad_norm": 0.3671875,
      "learning_rate": 1.6424440511052268e-05,
      "loss": 0.6431,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2575
    },
    {
      "epoch": 0.35028555887952134,
      "grad_norm": 0.310546875,
      "learning_rate": 1.642080142404966e-05,
      "loss": 0.6021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2576
    },
    {
      "epoch": 0.350421539298341,
      "grad_norm": 0.40625,
      "learning_rate": 1.6417160889719915e-05,
      "loss": 0.652,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2577
    },
    {
      "epoch": 0.3505575197171607,
      "grad_norm": 0.5546875,
      "learning_rate": 1.6413518908883646e-05,
      "loss": 0.735,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2578
    },
    {
      "epoch": 0.35069350013598044,
      "grad_norm": 0.47265625,
      "learning_rate": 1.6409875482361803e-05,
      "loss": 0.7581,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2579
    },
    {
      "epoch": 0.3508294805548001,
      "grad_norm": 0.345703125,
      "learning_rate": 1.6406230610975654e-05,
      "loss": 0.6017,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2580
    },
    {
      "epoch": 0.3509654609736198,
      "grad_norm": 0.41015625,
      "learning_rate": 1.6402584295546808e-05,
      "loss": 0.8162,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2581
    },
    {
      "epoch": 0.3511014413924395,
      "grad_norm": 0.26171875,
      "learning_rate": 1.6398936536897182e-05,
      "loss": 0.5345,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2582
    },
    {
      "epoch": 0.35123742181125917,
      "grad_norm": 0.462890625,
      "learning_rate": 1.6395287335849028e-05,
      "loss": 0.8127,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2583
    },
    {
      "epoch": 0.35137340223007885,
      "grad_norm": 0.3125,
      "learning_rate": 1.6391636693224916e-05,
      "loss": 0.5386,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2584
    },
    {
      "epoch": 0.3515093826488986,
      "grad_norm": 0.400390625,
      "learning_rate": 1.638798460984775e-05,
      "loss": 0.6961,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2585
    },
    {
      "epoch": 0.35164536306771826,
      "grad_norm": 0.53125,
      "learning_rate": 1.6384331086540753e-05,
      "loss": 0.4741,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2586
    },
    {
      "epoch": 0.35178134348653795,
      "grad_norm": 0.3984375,
      "learning_rate": 1.6380676124127472e-05,
      "loss": 0.6968,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2587
    },
    {
      "epoch": 0.3519173239053576,
      "grad_norm": 0.66015625,
      "learning_rate": 1.6377019723431783e-05,
      "loss": 0.6634,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2588
    },
    {
      "epoch": 0.3520533043241773,
      "grad_norm": 0.4375,
      "learning_rate": 1.6373361885277882e-05,
      "loss": 0.748,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2589
    },
    {
      "epoch": 0.352189284742997,
      "grad_norm": 0.193359375,
      "learning_rate": 1.6369702610490295e-05,
      "loss": 0.4435,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2590
    },
    {
      "epoch": 0.3523252651618167,
      "grad_norm": 0.35546875,
      "learning_rate": 1.636604189989386e-05,
      "loss": 0.7449,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2591
    },
    {
      "epoch": 0.3524612455806364,
      "grad_norm": 0.462890625,
      "learning_rate": 1.636237975431375e-05,
      "loss": 0.7464,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2592
    },
    {
      "epoch": 0.3525972259994561,
      "grad_norm": 0.8671875,
      "learning_rate": 1.6358716174575452e-05,
      "loss": 0.9137,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2593
    },
    {
      "epoch": 0.35273320641827577,
      "grad_norm": 0.30078125,
      "learning_rate": 1.635505116150479e-05,
      "loss": 0.4803,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2594
    },
    {
      "epoch": 0.35286918683709545,
      "grad_norm": 0.30859375,
      "learning_rate": 1.6351384715927897e-05,
      "loss": 0.5531,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2595
    },
    {
      "epoch": 0.35300516725591513,
      "grad_norm": 0.353515625,
      "learning_rate": 1.6347716838671238e-05,
      "loss": 0.5023,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2596
    },
    {
      "epoch": 0.3531411476747348,
      "grad_norm": 0.478515625,
      "learning_rate": 1.6344047530561597e-05,
      "loss": 0.849,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2597
    },
    {
      "epoch": 0.35327712809355455,
      "grad_norm": 0.53125,
      "learning_rate": 1.634037679242608e-05,
      "loss": 0.6446,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2598
    },
    {
      "epoch": 0.35341310851237423,
      "grad_norm": 0.3046875,
      "learning_rate": 1.6336704625092114e-05,
      "loss": 0.5228,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2599
    },
    {
      "epoch": 0.3535490889311939,
      "grad_norm": 0.2890625,
      "learning_rate": 1.633303102938745e-05,
      "loss": 0.6147,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2600
    },
    {
      "epoch": 0.3536850693500136,
      "grad_norm": 0.44140625,
      "learning_rate": 1.632935600614017e-05,
      "loss": 0.6817,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2601
    },
    {
      "epoch": 0.3538210497688333,
      "grad_norm": 0.5078125,
      "learning_rate": 1.6325679556178656e-05,
      "loss": 0.9048,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2602
    },
    {
      "epoch": 0.35395703018765295,
      "grad_norm": 0.357421875,
      "learning_rate": 1.6322001680331635e-05,
      "loss": 0.616,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2603
    },
    {
      "epoch": 0.3540930106064727,
      "grad_norm": 0.32421875,
      "learning_rate": 1.6318322379428144e-05,
      "loss": 0.6077,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2604
    },
    {
      "epoch": 0.3542289910252924,
      "grad_norm": 0.341796875,
      "learning_rate": 1.631464165429754e-05,
      "loss": 0.6692,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2605
    },
    {
      "epoch": 0.35436497144411205,
      "grad_norm": 0.361328125,
      "learning_rate": 1.6310959505769503e-05,
      "loss": 0.7266,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2606
    },
    {
      "epoch": 0.35450095186293173,
      "grad_norm": 0.546875,
      "learning_rate": 1.6307275934674034e-05,
      "loss": 0.7992,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2607
    },
    {
      "epoch": 0.3546369322817514,
      "grad_norm": 0.380859375,
      "learning_rate": 1.6303590941841457e-05,
      "loss": 0.6224,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2608
    },
    {
      "epoch": 0.3547729127005711,
      "grad_norm": 0.31640625,
      "learning_rate": 1.6299904528102414e-05,
      "loss": 0.6296,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2609
    },
    {
      "epoch": 0.35490889311939083,
      "grad_norm": 0.330078125,
      "learning_rate": 1.6296216694287864e-05,
      "loss": 0.6268,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2610
    },
    {
      "epoch": 0.3550448735382105,
      "grad_norm": 0.232421875,
      "learning_rate": 1.62925274412291e-05,
      "loss": 0.402,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2611
    },
    {
      "epoch": 0.3551808539570302,
      "grad_norm": 0.384765625,
      "learning_rate": 1.6288836769757712e-05,
      "loss": 0.563,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2612
    },
    {
      "epoch": 0.3553168343758499,
      "grad_norm": 0.328125,
      "learning_rate": 1.628514468070563e-05,
      "loss": 0.5687,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2613
    },
    {
      "epoch": 0.35545281479466956,
      "grad_norm": 0.357421875,
      "learning_rate": 1.6281451174905093e-05,
      "loss": 0.7492,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2614
    },
    {
      "epoch": 0.35558879521348924,
      "grad_norm": 1.4296875,
      "learning_rate": 1.6277756253188663e-05,
      "loss": 0.7646,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2615
    },
    {
      "epoch": 0.3557247756323089,
      "grad_norm": 0.4375,
      "learning_rate": 1.6274059916389223e-05,
      "loss": 0.4924,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2616
    },
    {
      "epoch": 0.35586075605112866,
      "grad_norm": 0.431640625,
      "learning_rate": 1.627036216533997e-05,
      "loss": 0.7158,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2617
    },
    {
      "epoch": 0.35599673646994834,
      "grad_norm": 0.73828125,
      "learning_rate": 1.6266663000874423e-05,
      "loss": 0.7209,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2618
    },
    {
      "epoch": 0.356132716888768,
      "grad_norm": 0.3984375,
      "learning_rate": 1.626296242382642e-05,
      "loss": 0.6336,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2619
    },
    {
      "epoch": 0.3562686973075877,
      "grad_norm": 0.6484375,
      "learning_rate": 1.6259260435030114e-05,
      "loss": 0.811,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2620
    },
    {
      "epoch": 0.3564046777264074,
      "grad_norm": 0.921875,
      "learning_rate": 1.625555703531998e-05,
      "loss": 0.75,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2621
    },
    {
      "epoch": 0.35654065814522706,
      "grad_norm": 0.875,
      "learning_rate": 1.625185222553081e-05,
      "loss": 0.6517,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2622
    },
    {
      "epoch": 0.3566766385640468,
      "grad_norm": 0.470703125,
      "learning_rate": 1.6248146006497712e-05,
      "loss": 0.4359,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2623
    },
    {
      "epoch": 0.3568126189828665,
      "grad_norm": 0.66015625,
      "learning_rate": 1.6244438379056115e-05,
      "loss": 0.6253,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2624
    },
    {
      "epoch": 0.35694859940168616,
      "grad_norm": 0.6484375,
      "learning_rate": 1.6240729344041762e-05,
      "loss": 0.783,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2625
    },
    {
      "epoch": 0.35708457982050584,
      "grad_norm": 0.40234375,
      "learning_rate": 1.6237018902290717e-05,
      "loss": 0.6021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2626
    },
    {
      "epoch": 0.3572205602393255,
      "grad_norm": 0.2412109375,
      "learning_rate": 1.6233307054639353e-05,
      "loss": 0.5015,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2627
    },
    {
      "epoch": 0.3573565406581452,
      "grad_norm": 0.298828125,
      "learning_rate": 1.622959380192437e-05,
      "loss": 0.3491,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2628
    },
    {
      "epoch": 0.35749252107696494,
      "grad_norm": 0.31640625,
      "learning_rate": 1.6225879144982783e-05,
      "loss": 0.5042,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2629
    },
    {
      "epoch": 0.3576285014957846,
      "grad_norm": 0.322265625,
      "learning_rate": 1.6222163084651918e-05,
      "loss": 0.6387,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2630
    },
    {
      "epoch": 0.3577644819146043,
      "grad_norm": 0.3984375,
      "learning_rate": 1.6218445621769424e-05,
      "loss": 0.6278,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2631
    },
    {
      "epoch": 0.357900462333424,
      "grad_norm": 0.65625,
      "learning_rate": 1.6214726757173255e-05,
      "loss": 0.8075,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2632
    },
    {
      "epoch": 0.35803644275224367,
      "grad_norm": 0.498046875,
      "learning_rate": 1.6211006491701695e-05,
      "loss": 0.5348,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2633
    },
    {
      "epoch": 0.35817242317106335,
      "grad_norm": 0.384765625,
      "learning_rate": 1.6207284826193334e-05,
      "loss": 0.5345,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2634
    },
    {
      "epoch": 0.3583084035898831,
      "grad_norm": 0.439453125,
      "learning_rate": 1.6203561761487084e-05,
      "loss": 0.6545,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2635
    },
    {
      "epoch": 0.35844438400870277,
      "grad_norm": 0.5625,
      "learning_rate": 1.6199837298422167e-05,
      "loss": 0.8231,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2636
    },
    {
      "epoch": 0.35858036442752245,
      "grad_norm": 0.4609375,
      "learning_rate": 1.6196111437838126e-05,
      "loss": 0.7222,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2637
    },
    {
      "epoch": 0.35871634484634213,
      "grad_norm": 0.40625,
      "learning_rate": 1.619238418057481e-05,
      "loss": 0.6375,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2638
    },
    {
      "epoch": 0.3588523252651618,
      "grad_norm": 0.466796875,
      "learning_rate": 1.618865552747239e-05,
      "loss": 0.6533,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2639
    },
    {
      "epoch": 0.3589883056839815,
      "grad_norm": 0.48046875,
      "learning_rate": 1.618492547937135e-05,
      "loss": 0.4797,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2640
    },
    {
      "epoch": 0.35912428610280117,
      "grad_norm": 0.609375,
      "learning_rate": 1.6181194037112495e-05,
      "loss": 0.6901,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2641
    },
    {
      "epoch": 0.3592602665216209,
      "grad_norm": 0.6484375,
      "learning_rate": 1.617746120153693e-05,
      "loss": 0.4272,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2642
    },
    {
      "epoch": 0.3593962469404406,
      "grad_norm": 0.58203125,
      "learning_rate": 1.6173726973486084e-05,
      "loss": 0.762,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2643
    },
    {
      "epoch": 0.35953222735926027,
      "grad_norm": 0.244140625,
      "learning_rate": 1.6169991353801696e-05,
      "loss": 0.4235,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2644
    },
    {
      "epoch": 0.35966820777807995,
      "grad_norm": 0.453125,
      "learning_rate": 1.6166254343325824e-05,
      "loss": 0.714,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2645
    },
    {
      "epoch": 0.35980418819689963,
      "grad_norm": 0.5546875,
      "learning_rate": 1.616251594290083e-05,
      "loss": 0.6821,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2646
    },
    {
      "epoch": 0.3599401686157193,
      "grad_norm": 0.55859375,
      "learning_rate": 1.6158776153369406e-05,
      "loss": 0.6506,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2647
    },
    {
      "epoch": 0.36007614903453905,
      "grad_norm": 0.384765625,
      "learning_rate": 1.6155034975574534e-05,
      "loss": 0.6559,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2648
    },
    {
      "epoch": 0.36021212945335873,
      "grad_norm": 0.3125,
      "learning_rate": 1.6151292410359528e-05,
      "loss": 0.5596,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2649
    },
    {
      "epoch": 0.3603481098721784,
      "grad_norm": 0.3203125,
      "learning_rate": 1.6147548458568007e-05,
      "loss": 0.5444,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2650
    },
    {
      "epoch": 0.3604840902909981,
      "grad_norm": 0.34375,
      "learning_rate": 1.6143803121043903e-05,
      "loss": 0.5524,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2651
    },
    {
      "epoch": 0.3606200707098178,
      "grad_norm": 0.376953125,
      "learning_rate": 1.614005639863146e-05,
      "loss": 0.6405,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2652
    },
    {
      "epoch": 0.36075605112863746,
      "grad_norm": 0.6484375,
      "learning_rate": 1.613630829217524e-05,
      "loss": 0.6126,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2653
    },
    {
      "epoch": 0.3608920315474572,
      "grad_norm": 0.859375,
      "learning_rate": 1.6132558802520103e-05,
      "loss": 0.5557,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2654
    },
    {
      "epoch": 0.3610280119662769,
      "grad_norm": 0.3671875,
      "learning_rate": 1.6128807930511235e-05,
      "loss": 0.6457,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2655
    },
    {
      "epoch": 0.36116399238509656,
      "grad_norm": 0.49609375,
      "learning_rate": 1.6125055676994132e-05,
      "loss": 0.8413,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2656
    },
    {
      "epoch": 0.36129997280391624,
      "grad_norm": 0.5078125,
      "learning_rate": 1.6121302042814588e-05,
      "loss": 0.797,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2657
    },
    {
      "epoch": 0.3614359532227359,
      "grad_norm": 0.404296875,
      "learning_rate": 1.611754702881873e-05,
      "loss": 0.6621,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2658
    },
    {
      "epoch": 0.3615719336415556,
      "grad_norm": 0.431640625,
      "learning_rate": 1.6113790635852974e-05,
      "loss": 0.6737,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2659
    },
    {
      "epoch": 0.3617079140603753,
      "grad_norm": 0.498046875,
      "learning_rate": 1.611003286476406e-05,
      "loss": 0.7634,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2660
    },
    {
      "epoch": 0.361843894479195,
      "grad_norm": 0.32421875,
      "learning_rate": 1.610627371639904e-05,
      "loss": 0.5285,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2661
    },
    {
      "epoch": 0.3619798748980147,
      "grad_norm": 0.330078125,
      "learning_rate": 1.610251319160527e-05,
      "loss": 0.5342,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2662
    },
    {
      "epoch": 0.3621158553168344,
      "grad_norm": 0.421875,
      "learning_rate": 1.6098751291230414e-05,
      "loss": 0.7539,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2663
    },
    {
      "epoch": 0.36225183573565406,
      "grad_norm": 0.3125,
      "learning_rate": 1.6094988016122454e-05,
      "loss": 0.6187,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2664
    },
    {
      "epoch": 0.36238781615447374,
      "grad_norm": 0.353515625,
      "learning_rate": 1.6091223367129676e-05,
      "loss": 0.6323,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2665
    },
    {
      "epoch": 0.3625237965732934,
      "grad_norm": 0.2578125,
      "learning_rate": 1.6087457345100678e-05,
      "loss": 0.4728,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2666
    },
    {
      "epoch": 0.36265977699211316,
      "grad_norm": 0.59375,
      "learning_rate": 1.608368995088437e-05,
      "loss": 0.7681,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2667
    },
    {
      "epoch": 0.36279575741093284,
      "grad_norm": 0.3203125,
      "learning_rate": 1.607992118532997e-05,
      "loss": 0.5472,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2668
    },
    {
      "epoch": 0.3629317378297525,
      "grad_norm": 0.310546875,
      "learning_rate": 1.6076151049287e-05,
      "loss": 0.5638,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2669
    },
    {
      "epoch": 0.3630677182485722,
      "grad_norm": 2.234375,
      "learning_rate": 1.6072379543605293e-05,
      "loss": 0.6829,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2670
    },
    {
      "epoch": 0.3632036986673919,
      "grad_norm": 0.5546875,
      "learning_rate": 1.6068606669135002e-05,
      "loss": 0.6213,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2671
    },
    {
      "epoch": 0.36333967908621156,
      "grad_norm": 0.265625,
      "learning_rate": 1.6064832426726566e-05,
      "loss": 0.5306,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2672
    },
    {
      "epoch": 0.3634756595050313,
      "grad_norm": 0.248046875,
      "learning_rate": 1.6061056817230754e-05,
      "loss": 0.4886,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2673
    },
    {
      "epoch": 0.363611639923851,
      "grad_norm": 0.31640625,
      "learning_rate": 1.6057279841498633e-05,
      "loss": 0.5646,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2674
    },
    {
      "epoch": 0.36374762034267066,
      "grad_norm": 0.482421875,
      "learning_rate": 1.605350150038158e-05,
      "loss": 0.5998,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2675
    },
    {
      "epoch": 0.36388360076149034,
      "grad_norm": 0.427734375,
      "learning_rate": 1.6049721794731276e-05,
      "loss": 0.3976,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2676
    },
    {
      "epoch": 0.36401958118031,
      "grad_norm": 0.59375,
      "learning_rate": 1.6045940725399715e-05,
      "loss": 0.5876,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2677
    },
    {
      "epoch": 0.3641555615991297,
      "grad_norm": 0.306640625,
      "learning_rate": 1.6042158293239197e-05,
      "loss": 0.5479,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2678
    },
    {
      "epoch": 0.3642915420179494,
      "grad_norm": 0.341796875,
      "learning_rate": 1.603837449910233e-05,
      "loss": 0.619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2679
    },
    {
      "epoch": 0.3644275224367691,
      "grad_norm": 0.68359375,
      "learning_rate": 1.603458934384202e-05,
      "loss": 0.4867,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2680
    },
    {
      "epoch": 0.3645635028555888,
      "grad_norm": 0.439453125,
      "learning_rate": 1.6030802828311498e-05,
      "loss": 0.6885,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2681
    },
    {
      "epoch": 0.3646994832744085,
      "grad_norm": 0.24609375,
      "learning_rate": 1.6027014953364282e-05,
      "loss": 0.438,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2682
    },
    {
      "epoch": 0.36483546369322817,
      "grad_norm": 0.435546875,
      "learning_rate": 1.602322571985421e-05,
      "loss": 0.6851,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2683
    },
    {
      "epoch": 0.36497144411204785,
      "grad_norm": 0.4609375,
      "learning_rate": 1.6019435128635424e-05,
      "loss": 0.7129,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2684
    },
    {
      "epoch": 0.36510742453086753,
      "grad_norm": 0.58984375,
      "learning_rate": 1.601564318056236e-05,
      "loss": 0.6169,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2685
    },
    {
      "epoch": 0.36524340494968727,
      "grad_norm": 0.5234375,
      "learning_rate": 1.6011849876489777e-05,
      "loss": 0.6216,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2686
    },
    {
      "epoch": 0.36537938536850695,
      "grad_norm": 0.498046875,
      "learning_rate": 1.6008055217272733e-05,
      "loss": 0.9095,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2687
    },
    {
      "epoch": 0.36551536578732663,
      "grad_norm": 0.271484375,
      "learning_rate": 1.6004259203766586e-05,
      "loss": 0.4692,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2688
    },
    {
      "epoch": 0.3656513462061463,
      "grad_norm": 0.6796875,
      "learning_rate": 1.6000461836827007e-05,
      "loss": 0.5601,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2689
    },
    {
      "epoch": 0.365787326624966,
      "grad_norm": 0.890625,
      "learning_rate": 1.5996663117309965e-05,
      "loss": 0.8424,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2690
    },
    {
      "epoch": 0.3659233070437857,
      "grad_norm": 0.3515625,
      "learning_rate": 1.5992863046071744e-05,
      "loss": 0.6379,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2691
    },
    {
      "epoch": 0.3660592874626054,
      "grad_norm": 0.3203125,
      "learning_rate": 1.598906162396892e-05,
      "loss": 0.6825,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2692
    },
    {
      "epoch": 0.3661952678814251,
      "grad_norm": 0.31640625,
      "learning_rate": 1.5985258851858385e-05,
      "loss": 0.4961,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2693
    },
    {
      "epoch": 0.36633124830024477,
      "grad_norm": 0.470703125,
      "learning_rate": 1.598145473059733e-05,
      "loss": 0.6844,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2694
    },
    {
      "epoch": 0.36646722871906445,
      "grad_norm": 0.9375,
      "learning_rate": 1.5977649261043247e-05,
      "loss": 0.694,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2695
    },
    {
      "epoch": 0.36660320913788413,
      "grad_norm": 0.58203125,
      "learning_rate": 1.5973842444053937e-05,
      "loss": 0.7096,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2696
    },
    {
      "epoch": 0.3667391895567038,
      "grad_norm": 0.4609375,
      "learning_rate": 1.5970034280487504e-05,
      "loss": 0.6794,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2697
    },
    {
      "epoch": 0.36687516997552355,
      "grad_norm": 0.41015625,
      "learning_rate": 1.5966224771202355e-05,
      "loss": 0.5661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2698
    },
    {
      "epoch": 0.36701115039434323,
      "grad_norm": 0.408203125,
      "learning_rate": 1.59624139170572e-05,
      "loss": 0.5386,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2699
    },
    {
      "epoch": 0.3671471308131629,
      "grad_norm": 0.5234375,
      "learning_rate": 1.5958601718911048e-05,
      "loss": 0.8781,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2700
    },
    {
      "epoch": 0.3672831112319826,
      "grad_norm": 0.5625,
      "learning_rate": 1.5954788177623218e-05,
      "loss": 0.6095,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2701
    },
    {
      "epoch": 0.3674190916508023,
      "grad_norm": 0.33203125,
      "learning_rate": 1.5950973294053332e-05,
      "loss": 0.5425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2702
    },
    {
      "epoch": 0.36755507206962196,
      "grad_norm": 0.4296875,
      "learning_rate": 1.594715706906131e-05,
      "loss": 0.597,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2703
    },
    {
      "epoch": 0.36769105248844164,
      "grad_norm": 0.494140625,
      "learning_rate": 1.5943339503507372e-05,
      "loss": 0.6323,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2704
    },
    {
      "epoch": 0.3678270329072614,
      "grad_norm": 0.37890625,
      "learning_rate": 1.5939520598252042e-05,
      "loss": 0.6242,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2705
    },
    {
      "epoch": 0.36796301332608106,
      "grad_norm": 0.65234375,
      "learning_rate": 1.5935700354156158e-05,
      "loss": 0.7591,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2706
    },
    {
      "epoch": 0.36809899374490074,
      "grad_norm": 0.2236328125,
      "learning_rate": 1.593187877208084e-05,
      "loss": 0.416,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2707
    },
    {
      "epoch": 0.3682349741637204,
      "grad_norm": 0.486328125,
      "learning_rate": 1.5928055852887525e-05,
      "loss": 0.5529,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2708
    },
    {
      "epoch": 0.3683709545825401,
      "grad_norm": 0.333984375,
      "learning_rate": 1.5924231597437946e-05,
      "loss": 0.7039,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2709
    },
    {
      "epoch": 0.3685069350013598,
      "grad_norm": 0.6484375,
      "learning_rate": 1.592040600659413e-05,
      "loss": 0.5706,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2710
    },
    {
      "epoch": 0.3686429154201795,
      "grad_norm": 0.412109375,
      "learning_rate": 1.591657908121842e-05,
      "loss": 0.6704,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2711
    },
    {
      "epoch": 0.3687788958389992,
      "grad_norm": 0.375,
      "learning_rate": 1.5912750822173446e-05,
      "loss": 0.7546,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2712
    },
    {
      "epoch": 0.3689148762578189,
      "grad_norm": 0.294921875,
      "learning_rate": 1.590892123032215e-05,
      "loss": 0.5669,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2713
    },
    {
      "epoch": 0.36905085667663856,
      "grad_norm": 0.3984375,
      "learning_rate": 1.5905090306527763e-05,
      "loss": 0.4676,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2714
    },
    {
      "epoch": 0.36918683709545824,
      "grad_norm": 0.330078125,
      "learning_rate": 1.5901258051653826e-05,
      "loss": 0.6615,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2715
    },
    {
      "epoch": 0.3693228175142779,
      "grad_norm": 0.349609375,
      "learning_rate": 1.5897424466564176e-05,
      "loss": 0.6366,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2716
    },
    {
      "epoch": 0.36945879793309766,
      "grad_norm": 0.423828125,
      "learning_rate": 1.589358955212295e-05,
      "loss": 0.7176,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2717
    },
    {
      "epoch": 0.36959477835191734,
      "grad_norm": 0.255859375,
      "learning_rate": 1.588975330919458e-05,
      "loss": 0.3519,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2718
    },
    {
      "epoch": 0.369730758770737,
      "grad_norm": 0.287109375,
      "learning_rate": 1.5885915738643807e-05,
      "loss": 0.5052,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2719
    },
    {
      "epoch": 0.3698667391895567,
      "grad_norm": 0.337890625,
      "learning_rate": 1.5882076841335664e-05,
      "loss": 0.6841,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2720
    },
    {
      "epoch": 0.3700027196083764,
      "grad_norm": 0.3984375,
      "learning_rate": 1.587823661813549e-05,
      "loss": 0.6639,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2721
    },
    {
      "epoch": 0.37013870002719607,
      "grad_norm": 0.310546875,
      "learning_rate": 1.587439506990891e-05,
      "loss": 0.6451,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2722
    },
    {
      "epoch": 0.37027468044601575,
      "grad_norm": 0.4765625,
      "learning_rate": 1.587055219752186e-05,
      "loss": 0.7825,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2723
    },
    {
      "epoch": 0.3704106608648355,
      "grad_norm": 0.294921875,
      "learning_rate": 1.5866708001840572e-05,
      "loss": 0.6424,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2724
    },
    {
      "epoch": 0.37054664128365516,
      "grad_norm": 0.4921875,
      "learning_rate": 1.5862862483731574e-05,
      "loss": 0.7075,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2725
    },
    {
      "epoch": 0.37068262170247485,
      "grad_norm": 0.431640625,
      "learning_rate": 1.5859015644061694e-05,
      "loss": 0.7773,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2726
    },
    {
      "epoch": 0.3708186021212945,
      "grad_norm": 0.41796875,
      "learning_rate": 1.5855167483698056e-05,
      "loss": 0.6624,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2727
    },
    {
      "epoch": 0.3709545825401142,
      "grad_norm": 0.54296875,
      "learning_rate": 1.585131800350808e-05,
      "loss": 0.6457,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2728
    },
    {
      "epoch": 0.3710905629589339,
      "grad_norm": 0.380859375,
      "learning_rate": 1.5847467204359482e-05,
      "loss": 0.5843,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2729
    },
    {
      "epoch": 0.3712265433777536,
      "grad_norm": 0.263671875,
      "learning_rate": 1.584361508712029e-05,
      "loss": 0.5452,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2730
    },
    {
      "epoch": 0.3713625237965733,
      "grad_norm": 0.69140625,
      "learning_rate": 1.5839761652658815e-05,
      "loss": 0.6134,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2731
    },
    {
      "epoch": 0.371498504215393,
      "grad_norm": 0.431640625,
      "learning_rate": 1.583590690184366e-05,
      "loss": 0.8175,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2732
    },
    {
      "epoch": 0.37163448463421267,
      "grad_norm": 0.58203125,
      "learning_rate": 1.5832050835543742e-05,
      "loss": 0.5243,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2733
    },
    {
      "epoch": 0.37177046505303235,
      "grad_norm": 0.37890625,
      "learning_rate": 1.582819345462826e-05,
      "loss": 0.5959,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2734
    },
    {
      "epoch": 0.37190644547185203,
      "grad_norm": 0.8359375,
      "learning_rate": 1.5824334759966725e-05,
      "loss": 0.7362,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2735
    },
    {
      "epoch": 0.37204242589067177,
      "grad_norm": 0.466796875,
      "learning_rate": 1.582047475242892e-05,
      "loss": 0.7212,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2736
    },
    {
      "epoch": 0.37217840630949145,
      "grad_norm": 0.30078125,
      "learning_rate": 1.5816613432884945e-05,
      "loss": 0.5247,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2737
    },
    {
      "epoch": 0.37231438672831113,
      "grad_norm": 0.50390625,
      "learning_rate": 1.5812750802205187e-05,
      "loss": 0.7292,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2738
    },
    {
      "epoch": 0.3724503671471308,
      "grad_norm": 0.65234375,
      "learning_rate": 1.580888686126033e-05,
      "loss": 0.5954,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2739
    },
    {
      "epoch": 0.3725863475659505,
      "grad_norm": 0.416015625,
      "learning_rate": 1.5805021610921357e-05,
      "loss": 0.7202,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2740
    },
    {
      "epoch": 0.3727223279847702,
      "grad_norm": 0.37109375,
      "learning_rate": 1.5801155052059543e-05,
      "loss": 0.6707,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2741
    },
    {
      "epoch": 0.37285830840358986,
      "grad_norm": 0.40625,
      "learning_rate": 1.579728718554645e-05,
      "loss": 0.549,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2742
    },
    {
      "epoch": 0.3729942888224096,
      "grad_norm": 0.474609375,
      "learning_rate": 1.579341801225395e-05,
      "loss": 0.6751,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2743
    },
    {
      "epoch": 0.3731302692412293,
      "grad_norm": 0.50390625,
      "learning_rate": 1.57895475330542e-05,
      "loss": 0.632,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2744
    },
    {
      "epoch": 0.37326624966004895,
      "grad_norm": 0.41015625,
      "learning_rate": 1.5785675748819652e-05,
      "loss": 0.6543,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2745
    },
    {
      "epoch": 0.37340223007886864,
      "grad_norm": 0.3828125,
      "learning_rate": 1.5781802660423055e-05,
      "loss": 0.6748,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2746
    },
    {
      "epoch": 0.3735382104976883,
      "grad_norm": 0.353515625,
      "learning_rate": 1.5777928268737442e-05,
      "loss": 0.7225,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2747
    },
    {
      "epoch": 0.373674190916508,
      "grad_norm": 0.68359375,
      "learning_rate": 1.5774052574636165e-05,
      "loss": 0.792,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2748
    },
    {
      "epoch": 0.37381017133532773,
      "grad_norm": 0.36328125,
      "learning_rate": 1.577017557899284e-05,
      "loss": 0.6462,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2749
    },
    {
      "epoch": 0.3739461517541474,
      "grad_norm": 0.6875,
      "learning_rate": 1.5766297282681393e-05,
      "loss": 0.7957,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2750
    },
    {
      "epoch": 0.3740821321729671,
      "grad_norm": 0.38671875,
      "learning_rate": 1.576241768657604e-05,
      "loss": 0.7246,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2751
    },
    {
      "epoch": 0.3742181125917868,
      "grad_norm": 0.37890625,
      "learning_rate": 1.5758536791551286e-05,
      "loss": 0.6642,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2752
    },
    {
      "epoch": 0.37435409301060646,
      "grad_norm": 0.3515625,
      "learning_rate": 1.5754654598481938e-05,
      "loss": 0.6519,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2753
    },
    {
      "epoch": 0.37449007342942614,
      "grad_norm": 0.53515625,
      "learning_rate": 1.5750771108243085e-05,
      "loss": 0.7554,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2754
    },
    {
      "epoch": 0.3746260538482459,
      "grad_norm": 0.26953125,
      "learning_rate": 1.574688632171012e-05,
      "loss": 0.5177,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2755
    },
    {
      "epoch": 0.37476203426706556,
      "grad_norm": 0.6875,
      "learning_rate": 1.5743000239758708e-05,
      "loss": 0.7083,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2756
    },
    {
      "epoch": 0.37489801468588524,
      "grad_norm": 0.66015625,
      "learning_rate": 1.573911286326483e-05,
      "loss": 0.752,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2757
    },
    {
      "epoch": 0.3750339951047049,
      "grad_norm": 0.345703125,
      "learning_rate": 1.573522419310475e-05,
      "loss": 0.6418,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2758
    },
    {
      "epoch": 0.3751699755235246,
      "grad_norm": 0.392578125,
      "learning_rate": 1.5731334230155012e-05,
      "loss": 0.7324,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2759
    },
    {
      "epoch": 0.3753059559423443,
      "grad_norm": 0.51171875,
      "learning_rate": 1.5727442975292468e-05,
      "loss": 0.5601,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2760
    },
    {
      "epoch": 0.375441936361164,
      "grad_norm": 0.33984375,
      "learning_rate": 1.5723550429394255e-05,
      "loss": 0.5544,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2761
    },
    {
      "epoch": 0.3755779167799837,
      "grad_norm": 0.310546875,
      "learning_rate": 1.57196565933378e-05,
      "loss": 0.6073,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2762
    },
    {
      "epoch": 0.3757138971988034,
      "grad_norm": 0.255859375,
      "learning_rate": 1.5715761468000817e-05,
      "loss": 0.5267,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2763
    },
    {
      "epoch": 0.37584987761762306,
      "grad_norm": 0.474609375,
      "learning_rate": 1.571186505426132e-05,
      "loss": 0.6419,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2764
    },
    {
      "epoch": 0.37598585803644274,
      "grad_norm": 0.431640625,
      "learning_rate": 1.5707967352997605e-05,
      "loss": 0.6359,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2765
    },
    {
      "epoch": 0.3761218384552624,
      "grad_norm": 0.78515625,
      "learning_rate": 1.570406836508826e-05,
      "loss": 0.8308,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2766
    },
    {
      "epoch": 0.3762578188740821,
      "grad_norm": 0.443359375,
      "learning_rate": 1.570016809141217e-05,
      "loss": 0.7744,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2767
    },
    {
      "epoch": 0.37639379929290184,
      "grad_norm": 0.443359375,
      "learning_rate": 1.56962665328485e-05,
      "loss": 0.7137,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2768
    },
    {
      "epoch": 0.3765297797117215,
      "grad_norm": 0.287109375,
      "learning_rate": 1.569236369027671e-05,
      "loss": 0.4831,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2769
    },
    {
      "epoch": 0.3766657601305412,
      "grad_norm": 0.55859375,
      "learning_rate": 1.568845956457655e-05,
      "loss": 0.5529,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2770
    },
    {
      "epoch": 0.3768017405493609,
      "grad_norm": 0.392578125,
      "learning_rate": 1.5684554156628057e-05,
      "loss": 0.5968,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2771
    },
    {
      "epoch": 0.37693772096818057,
      "grad_norm": 0.341796875,
      "learning_rate": 1.568064746731156e-05,
      "loss": 0.7044,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2772
    },
    {
      "epoch": 0.37707370138700025,
      "grad_norm": 0.349609375,
      "learning_rate": 1.5676739497507666e-05,
      "loss": 0.7349,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2773
    },
    {
      "epoch": 0.37720968180582,
      "grad_norm": 0.314453125,
      "learning_rate": 1.567283024809729e-05,
      "loss": 0.5596,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2774
    },
    {
      "epoch": 0.37734566222463967,
      "grad_norm": 0.62890625,
      "learning_rate": 1.5668919719961617e-05,
      "loss": 0.627,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2775
    },
    {
      "epoch": 0.37748164264345935,
      "grad_norm": 0.458984375,
      "learning_rate": 1.5665007913982132e-05,
      "loss": 0.6385,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2776
    },
    {
      "epoch": 0.37761762306227903,
      "grad_norm": 0.32421875,
      "learning_rate": 1.56610948310406e-05,
      "loss": 0.6159,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2777
    },
    {
      "epoch": 0.3777536034810987,
      "grad_norm": 0.3671875,
      "learning_rate": 1.5657180472019083e-05,
      "loss": 0.689,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2778
    },
    {
      "epoch": 0.3778895838999184,
      "grad_norm": 0.5546875,
      "learning_rate": 1.565326483779992e-05,
      "loss": 0.6722,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2779
    },
    {
      "epoch": 0.3780255643187381,
      "grad_norm": 0.41796875,
      "learning_rate": 1.5649347929265746e-05,
      "loss": 0.569,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2780
    },
    {
      "epoch": 0.3781615447375578,
      "grad_norm": 0.2734375,
      "learning_rate": 1.564542974729948e-05,
      "loss": 0.4167,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2781
    },
    {
      "epoch": 0.3782975251563775,
      "grad_norm": 0.4765625,
      "learning_rate": 1.5641510292784328e-05,
      "loss": 0.6797,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2782
    },
    {
      "epoch": 0.37843350557519717,
      "grad_norm": 0.3828125,
      "learning_rate": 1.563758956660378e-05,
      "loss": 0.7992,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2783
    },
    {
      "epoch": 0.37856948599401685,
      "grad_norm": 0.25390625,
      "learning_rate": 1.563366756964162e-05,
      "loss": 0.499,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2784
    },
    {
      "epoch": 0.37870546641283653,
      "grad_norm": 0.34765625,
      "learning_rate": 1.5629744302781913e-05,
      "loss": 0.665,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2785
    },
    {
      "epoch": 0.3788414468316562,
      "grad_norm": 0.33203125,
      "learning_rate": 1.5625819766909008e-05,
      "loss": 0.7259,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2786
    },
    {
      "epoch": 0.37897742725047595,
      "grad_norm": 0.55078125,
      "learning_rate": 1.5621893962907547e-05,
      "loss": 0.6278,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2787
    },
    {
      "epoch": 0.37911340766929563,
      "grad_norm": 0.369140625,
      "learning_rate": 1.561796689166245e-05,
      "loss": 0.6999,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2788
    },
    {
      "epoch": 0.3792493880881153,
      "grad_norm": 0.333984375,
      "learning_rate": 1.5614038554058936e-05,
      "loss": 0.5584,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2789
    },
    {
      "epoch": 0.379385368506935,
      "grad_norm": 0.5625,
      "learning_rate": 1.5610108950982494e-05,
      "loss": 0.7563,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2790
    },
    {
      "epoch": 0.3795213489257547,
      "grad_norm": 0.51171875,
      "learning_rate": 1.56061780833189e-05,
      "loss": 0.7319,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2791
    },
    {
      "epoch": 0.37965732934457436,
      "grad_norm": 0.330078125,
      "learning_rate": 1.5602245951954228e-05,
      "loss": 0.5832,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2792
    },
    {
      "epoch": 0.3797933097633941,
      "grad_norm": 0.314453125,
      "learning_rate": 1.5598312557774823e-05,
      "loss": 0.5213,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2793
    },
    {
      "epoch": 0.3799292901822138,
      "grad_norm": 0.345703125,
      "learning_rate": 1.5594377901667323e-05,
      "loss": 0.604,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2794
    },
    {
      "epoch": 0.38006527060103346,
      "grad_norm": 0.51171875,
      "learning_rate": 1.5590441984518644e-05,
      "loss": 0.3914,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2795
    },
    {
      "epoch": 0.38020125101985314,
      "grad_norm": 0.37109375,
      "learning_rate": 1.5586504807215997e-05,
      "loss": 0.641,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2796
    },
    {
      "epoch": 0.3803372314386728,
      "grad_norm": 0.30859375,
      "learning_rate": 1.558256637064686e-05,
      "loss": 0.491,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2797
    },
    {
      "epoch": 0.3804732118574925,
      "grad_norm": 0.423828125,
      "learning_rate": 1.5578626675699014e-05,
      "loss": 0.765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2798
    },
    {
      "epoch": 0.38060919227631224,
      "grad_norm": 0.53515625,
      "learning_rate": 1.557468572326051e-05,
      "loss": 0.535,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2799
    },
    {
      "epoch": 0.3807451726951319,
      "grad_norm": 0.5078125,
      "learning_rate": 1.5570743514219686e-05,
      "loss": 0.7161,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2800
    },
    {
      "epoch": 0.3808811531139516,
      "grad_norm": 0.3203125,
      "learning_rate": 1.5566800049465167e-05,
      "loss": 0.6234,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2801
    },
    {
      "epoch": 0.3810171335327713,
      "grad_norm": 0.51953125,
      "learning_rate": 1.5562855329885858e-05,
      "loss": 0.7213,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2802
    },
    {
      "epoch": 0.38115311395159096,
      "grad_norm": 0.77734375,
      "learning_rate": 1.5558909356370944e-05,
      "loss": 0.7332,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2803
    },
    {
      "epoch": 0.38128909437041064,
      "grad_norm": 0.4609375,
      "learning_rate": 1.5554962129809895e-05,
      "loss": 0.6341,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2804
    },
    {
      "epoch": 0.3814250747892304,
      "grad_norm": 0.50390625,
      "learning_rate": 1.555101365109247e-05,
      "loss": 0.7352,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2805
    },
    {
      "epoch": 0.38156105520805006,
      "grad_norm": 0.380859375,
      "learning_rate": 1.55470639211087e-05,
      "loss": 0.6245,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2806
    },
    {
      "epoch": 0.38169703562686974,
      "grad_norm": 0.53515625,
      "learning_rate": 1.5543112940748906e-05,
      "loss": 0.742,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2807
    },
    {
      "epoch": 0.3818330160456894,
      "grad_norm": 0.451171875,
      "learning_rate": 1.553916071090368e-05,
      "loss": 0.5397,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2808
    },
    {
      "epoch": 0.3819689964645091,
      "grad_norm": 0.3671875,
      "learning_rate": 1.553520723246391e-05,
      "loss": 0.6782,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2809
    },
    {
      "epoch": 0.3821049768833288,
      "grad_norm": 0.8359375,
      "learning_rate": 1.5531252506320755e-05,
      "loss": 0.6127,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2810
    },
    {
      "epoch": 0.38224095730214847,
      "grad_norm": 0.2578125,
      "learning_rate": 1.552729653336566e-05,
      "loss": 0.437,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2811
    },
    {
      "epoch": 0.3823769377209682,
      "grad_norm": 0.255859375,
      "learning_rate": 1.5523339314490348e-05,
      "loss": 0.5055,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2812
    },
    {
      "epoch": 0.3825129181397879,
      "grad_norm": 0.279296875,
      "learning_rate": 1.5519380850586828e-05,
      "loss": 0.5667,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2813
    },
    {
      "epoch": 0.38264889855860756,
      "grad_norm": 0.69140625,
      "learning_rate": 1.5515421142547383e-05,
      "loss": 0.7012,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2814
    },
    {
      "epoch": 0.38278487897742725,
      "grad_norm": 0.423828125,
      "learning_rate": 1.551146019126458e-05,
      "loss": 0.7304,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2815
    },
    {
      "epoch": 0.3829208593962469,
      "grad_norm": 0.341796875,
      "learning_rate": 1.5507497997631267e-05,
      "loss": 0.5103,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2816
    },
    {
      "epoch": 0.3830568398150666,
      "grad_norm": 0.37109375,
      "learning_rate": 1.5503534562540572e-05,
      "loss": 0.6278,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2817
    },
    {
      "epoch": 0.38319282023388634,
      "grad_norm": 0.3671875,
      "learning_rate": 1.5499569886885897e-05,
      "loss": 0.6901,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2818
    },
    {
      "epoch": 0.383328800652706,
      "grad_norm": 0.61328125,
      "learning_rate": 1.5495603971560932e-05,
      "loss": 0.9051,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2819
    },
    {
      "epoch": 0.3834647810715257,
      "grad_norm": 0.33203125,
      "learning_rate": 1.549163681745964e-05,
      "loss": 0.7305,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2820
    },
    {
      "epoch": 0.3836007614903454,
      "grad_norm": 0.314453125,
      "learning_rate": 1.548766842547627e-05,
      "loss": 0.5329,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2821
    },
    {
      "epoch": 0.38373674190916507,
      "grad_norm": 0.58984375,
      "learning_rate": 1.5483698796505345e-05,
      "loss": 0.5754,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2822
    },
    {
      "epoch": 0.38387272232798475,
      "grad_norm": 0.625,
      "learning_rate": 1.547972793144167e-05,
      "loss": 0.7661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2823
    },
    {
      "epoch": 0.3840087027468045,
      "grad_norm": 0.72265625,
      "learning_rate": 1.5475755831180318e-05,
      "loss": 0.5914,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2824
    },
    {
      "epoch": 0.38414468316562417,
      "grad_norm": 0.515625,
      "learning_rate": 1.5471782496616658e-05,
      "loss": 0.7695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2825
    },
    {
      "epoch": 0.38428066358444385,
      "grad_norm": 1.078125,
      "learning_rate": 1.5467807928646324e-05,
      "loss": 0.5741,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2826
    },
    {
      "epoch": 0.38441664400326353,
      "grad_norm": 0.435546875,
      "learning_rate": 1.546383212816523e-05,
      "loss": 0.7007,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2827
    },
    {
      "epoch": 0.3845526244220832,
      "grad_norm": 0.5,
      "learning_rate": 1.5459855096069577e-05,
      "loss": 0.6608,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2828
    },
    {
      "epoch": 0.3846886048409029,
      "grad_norm": 0.35546875,
      "learning_rate": 1.545587683325583e-05,
      "loss": 0.5877,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2829
    },
    {
      "epoch": 0.3848245852597226,
      "grad_norm": 0.47265625,
      "learning_rate": 1.5451897340620746e-05,
      "loss": 0.7152,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2830
    },
    {
      "epoch": 0.3849605656785423,
      "grad_norm": 0.27734375,
      "learning_rate": 1.544791661906134e-05,
      "loss": 0.5171,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2831
    },
    {
      "epoch": 0.385096546097362,
      "grad_norm": 0.76171875,
      "learning_rate": 1.5443934669474922e-05,
      "loss": 0.5845,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2832
    },
    {
      "epoch": 0.3852325265161817,
      "grad_norm": 0.4140625,
      "learning_rate": 1.5439951492759076e-05,
      "loss": 0.7458,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2833
    },
    {
      "epoch": 0.38536850693500135,
      "grad_norm": 0.3984375,
      "learning_rate": 1.543596708981165e-05,
      "loss": 0.6822,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2834
    },
    {
      "epoch": 0.38550448735382103,
      "grad_norm": 0.5546875,
      "learning_rate": 1.5431981461530786e-05,
      "loss": 0.5365,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2835
    },
    {
      "epoch": 0.3856404677726407,
      "grad_norm": 0.392578125,
      "learning_rate": 1.5427994608814884e-05,
      "loss": 0.6619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2836
    },
    {
      "epoch": 0.38577644819146045,
      "grad_norm": 0.3359375,
      "learning_rate": 1.5424006532562638e-05,
      "loss": 0.5345,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2837
    },
    {
      "epoch": 0.38591242861028013,
      "grad_norm": 0.29296875,
      "learning_rate": 1.5420017233673e-05,
      "loss": 0.4639,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2838
    },
    {
      "epoch": 0.3860484090290998,
      "grad_norm": 0.27734375,
      "learning_rate": 1.5416026713045217e-05,
      "loss": 0.4899,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2839
    },
    {
      "epoch": 0.3861843894479195,
      "grad_norm": 0.625,
      "learning_rate": 1.5412034971578792e-05,
      "loss": 0.5977,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2840
    },
    {
      "epoch": 0.3863203698667392,
      "grad_norm": 0.451171875,
      "learning_rate": 1.540804201017352e-05,
      "loss": 0.7557,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2841
    },
    {
      "epoch": 0.38645635028555886,
      "grad_norm": 0.35546875,
      "learning_rate": 1.540404782972946e-05,
      "loss": 0.6324,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2842
    },
    {
      "epoch": 0.3865923307043786,
      "grad_norm": 0.2412109375,
      "learning_rate": 1.5400052431146947e-05,
      "loss": 0.4434,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2843
    },
    {
      "epoch": 0.3867283111231983,
      "grad_norm": 0.60546875,
      "learning_rate": 1.5396055815326593e-05,
      "loss": 0.6156,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2844
    },
    {
      "epoch": 0.38686429154201796,
      "grad_norm": 0.392578125,
      "learning_rate": 1.539205798316929e-05,
      "loss": 0.7666,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2845
    },
    {
      "epoch": 0.38700027196083764,
      "grad_norm": 0.53515625,
      "learning_rate": 1.538805893557619e-05,
      "loss": 0.798,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2846
    },
    {
      "epoch": 0.3871362523796573,
      "grad_norm": 0.63671875,
      "learning_rate": 1.5384058673448733e-05,
      "loss": 0.786,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2847
    },
    {
      "epoch": 0.387272232798477,
      "grad_norm": 0.384765625,
      "learning_rate": 1.5380057197688624e-05,
      "loss": 0.7054,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2848
    },
    {
      "epoch": 0.3874082132172967,
      "grad_norm": 0.42578125,
      "learning_rate": 1.5376054509197845e-05,
      "loss": 0.6794,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2849
    },
    {
      "epoch": 0.3875441936361164,
      "grad_norm": 0.455078125,
      "learning_rate": 1.537205060887865e-05,
      "loss": 0.8424,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2850
    },
    {
      "epoch": 0.3876801740549361,
      "grad_norm": 0.345703125,
      "learning_rate": 1.5368045497633573e-05,
      "loss": 0.7103,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2851
    },
    {
      "epoch": 0.3878161544737558,
      "grad_norm": 0.3671875,
      "learning_rate": 1.536403917636541e-05,
      "loss": 0.5832,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2852
    },
    {
      "epoch": 0.38795213489257546,
      "grad_norm": 0.310546875,
      "learning_rate": 1.5360031645977236e-05,
      "loss": 0.6273,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2853
    },
    {
      "epoch": 0.38808811531139514,
      "grad_norm": 0.451171875,
      "learning_rate": 1.5356022907372393e-05,
      "loss": 0.6196,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2854
    },
    {
      "epoch": 0.3882240957302148,
      "grad_norm": 0.451171875,
      "learning_rate": 1.535201296145451e-05,
      "loss": 0.6683,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2855
    },
    {
      "epoch": 0.38836007614903456,
      "grad_norm": 0.474609375,
      "learning_rate": 1.5348001809127463e-05,
      "loss": 0.8644,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2856
    },
    {
      "epoch": 0.38849605656785424,
      "grad_norm": 0.35546875,
      "learning_rate": 1.5343989451295428e-05,
      "loss": 0.6172,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2857
    },
    {
      "epoch": 0.3886320369866739,
      "grad_norm": 0.455078125,
      "learning_rate": 1.5339975888862833e-05,
      "loss": 0.4321,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2858
    },
    {
      "epoch": 0.3887680174054936,
      "grad_norm": 0.26171875,
      "learning_rate": 1.533596112273439e-05,
      "loss": 0.5308,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2859
    },
    {
      "epoch": 0.3889039978243133,
      "grad_norm": 0.3828125,
      "learning_rate": 1.5331945153815065e-05,
      "loss": 0.7323,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2860
    },
    {
      "epoch": 0.38903997824313297,
      "grad_norm": 0.255859375,
      "learning_rate": 1.5327927983010118e-05,
      "loss": 0.5588,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2861
    },
    {
      "epoch": 0.3891759586619527,
      "grad_norm": 0.337890625,
      "learning_rate": 1.5323909611225062e-05,
      "loss": 0.5674,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2862
    },
    {
      "epoch": 0.3893119390807724,
      "grad_norm": 0.357421875,
      "learning_rate": 1.531989003936569e-05,
      "loss": 0.5996,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2863
    },
    {
      "epoch": 0.38944791949959207,
      "grad_norm": 0.8203125,
      "learning_rate": 1.5315869268338064e-05,
      "loss": 0.8621,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2864
    },
    {
      "epoch": 0.38958389991841175,
      "grad_norm": 0.287109375,
      "learning_rate": 1.531184729904851e-05,
      "loss": 0.6279,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2865
    },
    {
      "epoch": 0.38971988033723143,
      "grad_norm": 0.2265625,
      "learning_rate": 1.530782413240363e-05,
      "loss": 0.4801,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2866
    },
    {
      "epoch": 0.3898558607560511,
      "grad_norm": 0.345703125,
      "learning_rate": 1.53037997693103e-05,
      "loss": 0.617,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2867
    },
    {
      "epoch": 0.38999184117487085,
      "grad_norm": 0.287109375,
      "learning_rate": 1.529977421067566e-05,
      "loss": 0.6042,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2868
    },
    {
      "epoch": 0.3901278215936905,
      "grad_norm": 0.39453125,
      "learning_rate": 1.5295747457407114e-05,
      "loss": 0.7373,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2869
    },
    {
      "epoch": 0.3902638020125102,
      "grad_norm": 0.33984375,
      "learning_rate": 1.529171951041234e-05,
      "loss": 0.6453,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2870
    },
    {
      "epoch": 0.3903997824313299,
      "grad_norm": 0.51953125,
      "learning_rate": 1.52876903705993e-05,
      "loss": 0.577,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2871
    },
    {
      "epoch": 0.39053576285014957,
      "grad_norm": 0.41796875,
      "learning_rate": 1.5283660038876198e-05,
      "loss": 0.7435,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2872
    },
    {
      "epoch": 0.39067174326896925,
      "grad_norm": 0.333984375,
      "learning_rate": 1.5279628516151522e-05,
      "loss": 0.54,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2873
    },
    {
      "epoch": 0.39080772368778893,
      "grad_norm": 0.5,
      "learning_rate": 1.5275595803334036e-05,
      "loss": 0.5837,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2874
    },
    {
      "epoch": 0.39094370410660867,
      "grad_norm": 0.45703125,
      "learning_rate": 1.527156190133275e-05,
      "loss": 0.6595,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2875
    },
    {
      "epoch": 0.39107968452542835,
      "grad_norm": 0.47265625,
      "learning_rate": 1.5267526811056965e-05,
      "loss": 0.4061,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2876
    },
    {
      "epoch": 0.39121566494424803,
      "grad_norm": 0.34765625,
      "learning_rate": 1.5263490533416234e-05,
      "loss": 0.5885,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2877
    },
    {
      "epoch": 0.3913516453630677,
      "grad_norm": 0.400390625,
      "learning_rate": 1.5259453069320387e-05,
      "loss": 0.6484,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2878
    },
    {
      "epoch": 0.3914876257818874,
      "grad_norm": 0.3125,
      "learning_rate": 1.5255414419679512e-05,
      "loss": 0.6519,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2879
    },
    {
      "epoch": 0.3916236062007071,
      "grad_norm": 0.58203125,
      "learning_rate": 1.5251374585403977e-05,
      "loss": 0.8499,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2880
    },
    {
      "epoch": 0.3917595866195268,
      "grad_norm": 0.326171875,
      "learning_rate": 1.5247333567404407e-05,
      "loss": 0.6917,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2881
    },
    {
      "epoch": 0.3918955670383465,
      "grad_norm": 0.388671875,
      "learning_rate": 1.5243291366591693e-05,
      "loss": 0.6522,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2882
    },
    {
      "epoch": 0.3920315474571662,
      "grad_norm": 0.34375,
      "learning_rate": 1.5239247983877003e-05,
      "loss": 0.5412,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2883
    },
    {
      "epoch": 0.39216752787598586,
      "grad_norm": 0.55078125,
      "learning_rate": 1.523520342017176e-05,
      "loss": 0.576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2884
    },
    {
      "epoch": 0.39230350829480554,
      "grad_norm": 0.353515625,
      "learning_rate": 1.5231157676387662e-05,
      "loss": 0.6154,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2885
    },
    {
      "epoch": 0.3924394887136252,
      "grad_norm": 0.64453125,
      "learning_rate": 1.5227110753436666e-05,
      "loss": 0.7397,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2886
    },
    {
      "epoch": 0.39257546913244495,
      "grad_norm": 0.291015625,
      "learning_rate": 1.5223062652231e-05,
      "loss": 0.4572,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2887
    },
    {
      "epoch": 0.39271144955126464,
      "grad_norm": 0.3515625,
      "learning_rate": 1.5219013373683154e-05,
      "loss": 0.63,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2888
    },
    {
      "epoch": 0.3928474299700843,
      "grad_norm": 0.5390625,
      "learning_rate": 1.5214962918705884e-05,
      "loss": 0.7394,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2889
    },
    {
      "epoch": 0.392983410388904,
      "grad_norm": 0.2490234375,
      "learning_rate": 1.5210911288212215e-05,
      "loss": 0.4793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2890
    },
    {
      "epoch": 0.3931193908077237,
      "grad_norm": 0.412109375,
      "learning_rate": 1.5206858483115431e-05,
      "loss": 0.6047,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2891
    },
    {
      "epoch": 0.39325537122654336,
      "grad_norm": 0.427734375,
      "learning_rate": 1.5202804504329089e-05,
      "loss": 0.6011,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2892
    },
    {
      "epoch": 0.39339135164536304,
      "grad_norm": 0.4140625,
      "learning_rate": 1.5198749352767e-05,
      "loss": 0.7746,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2893
    },
    {
      "epoch": 0.3935273320641828,
      "grad_norm": 0.251953125,
      "learning_rate": 1.5194693029343249e-05,
      "loss": 0.3908,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2894
    },
    {
      "epoch": 0.39366331248300246,
      "grad_norm": 0.62890625,
      "learning_rate": 1.5190635534972173e-05,
      "loss": 0.8187,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2895
    },
    {
      "epoch": 0.39379929290182214,
      "grad_norm": 0.330078125,
      "learning_rate": 1.518657687056839e-05,
      "loss": 0.5902,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2896
    },
    {
      "epoch": 0.3939352733206418,
      "grad_norm": 0.310546875,
      "learning_rate": 1.5182517037046767e-05,
      "loss": 0.637,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2897
    },
    {
      "epoch": 0.3940712537394615,
      "grad_norm": 0.6328125,
      "learning_rate": 1.5178456035322445e-05,
      "loss": 0.4933,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2898
    },
    {
      "epoch": 0.3942072341582812,
      "grad_norm": 0.546875,
      "learning_rate": 1.517439386631082e-05,
      "loss": 0.6421,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2899
    },
    {
      "epoch": 0.3943432145771009,
      "grad_norm": 0.390625,
      "learning_rate": 1.5170330530927553e-05,
      "loss": 0.6299,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2900
    },
    {
      "epoch": 0.3944791949959206,
      "grad_norm": 0.63671875,
      "learning_rate": 1.516626603008857e-05,
      "loss": 0.6479,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2901
    },
    {
      "epoch": 0.3946151754147403,
      "grad_norm": 0.287109375,
      "learning_rate": 1.5162200364710065e-05,
      "loss": 0.5446,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2902
    },
    {
      "epoch": 0.39475115583355996,
      "grad_norm": 0.470703125,
      "learning_rate": 1.515813353570848e-05,
      "loss": 0.6864,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2903
    },
    {
      "epoch": 0.39488713625237964,
      "grad_norm": 0.40234375,
      "learning_rate": 1.5154065544000536e-05,
      "loss": 0.6665,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2904
    },
    {
      "epoch": 0.3950231166711993,
      "grad_norm": 0.341796875,
      "learning_rate": 1.5149996390503202e-05,
      "loss": 0.5425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2905
    },
    {
      "epoch": 0.39515909709001906,
      "grad_norm": 0.55859375,
      "learning_rate": 1.5145926076133716e-05,
      "loss": 0.7397,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2906
    },
    {
      "epoch": 0.39529507750883874,
      "grad_norm": 0.56640625,
      "learning_rate": 1.5141854601809583e-05,
      "loss": 0.6097,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2907
    },
    {
      "epoch": 0.3954310579276584,
      "grad_norm": 0.453125,
      "learning_rate": 1.5137781968448553e-05,
      "loss": 0.6382,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2908
    },
    {
      "epoch": 0.3955670383464781,
      "grad_norm": 0.33203125,
      "learning_rate": 1.5133708176968656e-05,
      "loss": 0.5864,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2909
    },
    {
      "epoch": 0.3957030187652978,
      "grad_norm": 0.5390625,
      "learning_rate": 1.5129633228288168e-05,
      "loss": 0.6488,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2910
    },
    {
      "epoch": 0.39583899918411747,
      "grad_norm": 1.1953125,
      "learning_rate": 1.5125557123325636e-05,
      "loss": 0.7477,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2911
    },
    {
      "epoch": 0.39597497960293715,
      "grad_norm": 0.326171875,
      "learning_rate": 1.5121479862999864e-05,
      "loss": 0.6369,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2912
    },
    {
      "epoch": 0.3961109600217569,
      "grad_norm": 0.4140625,
      "learning_rate": 1.5117401448229911e-05,
      "loss": 0.5926,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2913
    },
    {
      "epoch": 0.39624694044057657,
      "grad_norm": 0.35546875,
      "learning_rate": 1.511332187993511e-05,
      "loss": 0.6426,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2914
    },
    {
      "epoch": 0.39638292085939625,
      "grad_norm": 0.423828125,
      "learning_rate": 1.5109241159035042e-05,
      "loss": 0.8258,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2915
    },
    {
      "epoch": 0.39651890127821593,
      "grad_norm": 0.34375,
      "learning_rate": 1.5105159286449548e-05,
      "loss": 0.7085,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2916
    },
    {
      "epoch": 0.3966548816970356,
      "grad_norm": 0.34375,
      "learning_rate": 1.5101076263098738e-05,
      "loss": 0.6624,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2917
    },
    {
      "epoch": 0.3967908621158553,
      "grad_norm": 0.404296875,
      "learning_rate": 1.5096992089902969e-05,
      "loss": 0.6989,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2918
    },
    {
      "epoch": 0.39692684253467503,
      "grad_norm": 0.310546875,
      "learning_rate": 1.5092906767782868e-05,
      "loss": 0.4653,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2919
    },
    {
      "epoch": 0.3970628229534947,
      "grad_norm": 0.66796875,
      "learning_rate": 1.5088820297659314e-05,
      "loss": 0.5485,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2920
    },
    {
      "epoch": 0.3971988033723144,
      "grad_norm": 0.369140625,
      "learning_rate": 1.5084732680453451e-05,
      "loss": 0.7064,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2921
    },
    {
      "epoch": 0.39733478379113407,
      "grad_norm": 0.31640625,
      "learning_rate": 1.5080643917086675e-05,
      "loss": 0.5934,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2922
    },
    {
      "epoch": 0.39747076420995375,
      "grad_norm": 0.48046875,
      "learning_rate": 1.5076554008480641e-05,
      "loss": 0.8242,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2923
    },
    {
      "epoch": 0.39760674462877343,
      "grad_norm": 0.26953125,
      "learning_rate": 1.507246295555727e-05,
      "loss": 0.5749,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2924
    },
    {
      "epoch": 0.39774272504759317,
      "grad_norm": 0.70703125,
      "learning_rate": 1.506837075923873e-05,
      "loss": 0.8934,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2925
    },
    {
      "epoch": 0.39787870546641285,
      "grad_norm": 0.494140625,
      "learning_rate": 1.5064277420447458e-05,
      "loss": 0.4928,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2926
    },
    {
      "epoch": 0.39801468588523253,
      "grad_norm": 0.236328125,
      "learning_rate": 1.5060182940106138e-05,
      "loss": 0.4823,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2927
    },
    {
      "epoch": 0.3981506663040522,
      "grad_norm": 0.455078125,
      "learning_rate": 1.5056087319137716e-05,
      "loss": 0.5926,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2928
    },
    {
      "epoch": 0.3982866467228719,
      "grad_norm": 0.76171875,
      "learning_rate": 1.5051990558465397e-05,
      "loss": 0.7873,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2929
    },
    {
      "epoch": 0.3984226271416916,
      "grad_norm": 0.375,
      "learning_rate": 1.504789265901264e-05,
      "loss": 0.6271,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2930
    },
    {
      "epoch": 0.3985586075605113,
      "grad_norm": 0.298828125,
      "learning_rate": 1.504379362170316e-05,
      "loss": 0.5999,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2931
    },
    {
      "epoch": 0.398694587979331,
      "grad_norm": 0.306640625,
      "learning_rate": 1.5039693447460933e-05,
      "loss": 0.6229,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2932
    },
    {
      "epoch": 0.3988305683981507,
      "grad_norm": 0.359375,
      "learning_rate": 1.5035592137210188e-05,
      "loss": 0.5934,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2933
    },
    {
      "epoch": 0.39896654881697036,
      "grad_norm": 0.57421875,
      "learning_rate": 1.503148969187541e-05,
      "loss": 0.7352,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2934
    },
    {
      "epoch": 0.39910252923579004,
      "grad_norm": 0.333984375,
      "learning_rate": 1.502738611238134e-05,
      "loss": 0.6274,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2935
    },
    {
      "epoch": 0.3992385096546097,
      "grad_norm": 0.326171875,
      "learning_rate": 1.502328139965297e-05,
      "loss": 0.4159,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2936
    },
    {
      "epoch": 0.3993744900734294,
      "grad_norm": 0.1796875,
      "learning_rate": 1.5019175554615564e-05,
      "loss": 0.3363,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2937
    },
    {
      "epoch": 0.39951047049224914,
      "grad_norm": 0.69921875,
      "learning_rate": 1.501506857819462e-05,
      "loss": 0.6636,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2938
    },
    {
      "epoch": 0.3996464509110688,
      "grad_norm": 0.37890625,
      "learning_rate": 1.5010960471315905e-05,
      "loss": 0.813,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2939
    },
    {
      "epoch": 0.3997824313298885,
      "grad_norm": 0.392578125,
      "learning_rate": 1.5006851234905434e-05,
      "loss": 0.6592,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2940
    },
    {
      "epoch": 0.3999184117487082,
      "grad_norm": 0.412109375,
      "learning_rate": 1.5002740869889483e-05,
      "loss": 0.7186,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2941
    },
    {
      "epoch": 0.40005439216752786,
      "grad_norm": 0.453125,
      "learning_rate": 1.4998629377194576e-05,
      "loss": 0.6761,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2942
    },
    {
      "epoch": 0.40019037258634754,
      "grad_norm": 0.44140625,
      "learning_rate": 1.4994516757747491e-05,
      "loss": 0.6712,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2943
    },
    {
      "epoch": 0.4003263530051673,
      "grad_norm": 0.322265625,
      "learning_rate": 1.4990403012475265e-05,
      "loss": 0.5887,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2944
    },
    {
      "epoch": 0.40046233342398696,
      "grad_norm": 0.427734375,
      "learning_rate": 1.498628814230519e-05,
      "loss": 0.6826,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2945
    },
    {
      "epoch": 0.40059831384280664,
      "grad_norm": 0.498046875,
      "learning_rate": 1.4982172148164804e-05,
      "loss": 0.634,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2946
    },
    {
      "epoch": 0.4007342942616263,
      "grad_norm": 0.5,
      "learning_rate": 1.4978055030981902e-05,
      "loss": 0.708,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2947
    },
    {
      "epoch": 0.400870274680446,
      "grad_norm": 0.2392578125,
      "learning_rate": 1.4973936791684534e-05,
      "loss": 0.373,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2948
    },
    {
      "epoch": 0.4010062550992657,
      "grad_norm": 0.75,
      "learning_rate": 1.4969817431201002e-05,
      "loss": 0.6379,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2949
    },
    {
      "epoch": 0.4011422355180854,
      "grad_norm": 0.212890625,
      "learning_rate": 1.496569695045986e-05,
      "loss": 0.3936,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2950
    },
    {
      "epoch": 0.4012782159369051,
      "grad_norm": 0.408203125,
      "learning_rate": 1.4961575350389914e-05,
      "loss": 0.6401,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2951
    },
    {
      "epoch": 0.4014141963557248,
      "grad_norm": 0.64453125,
      "learning_rate": 1.4957452631920224e-05,
      "loss": 0.7842,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2952
    },
    {
      "epoch": 0.40155017677454447,
      "grad_norm": 0.41796875,
      "learning_rate": 1.49533287959801e-05,
      "loss": 0.6755,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2953
    },
    {
      "epoch": 0.40168615719336415,
      "grad_norm": 0.458984375,
      "learning_rate": 1.4949203843499108e-05,
      "loss": 0.5465,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2954
    },
    {
      "epoch": 0.4018221376121838,
      "grad_norm": 0.275390625,
      "learning_rate": 1.4945077775407058e-05,
      "loss": 0.4354,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2955
    },
    {
      "epoch": 0.4019581180310035,
      "grad_norm": 0.4296875,
      "learning_rate": 1.4940950592634018e-05,
      "loss": 0.5848,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2956
    },
    {
      "epoch": 0.40209409844982325,
      "grad_norm": 0.306640625,
      "learning_rate": 1.4936822296110307e-05,
      "loss": 0.5726,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2957
    },
    {
      "epoch": 0.4022300788686429,
      "grad_norm": 0.400390625,
      "learning_rate": 1.493269288676649e-05,
      "loss": 0.7619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2958
    },
    {
      "epoch": 0.4023660592874626,
      "grad_norm": 0.337890625,
      "learning_rate": 1.4928562365533393e-05,
      "loss": 0.6125,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2959
    },
    {
      "epoch": 0.4025020397062823,
      "grad_norm": 0.421875,
      "learning_rate": 1.4924430733342079e-05,
      "loss": 0.6675,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2960
    },
    {
      "epoch": 0.40263802012510197,
      "grad_norm": 0.56640625,
      "learning_rate": 1.492029799112387e-05,
      "loss": 0.6558,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2961
    },
    {
      "epoch": 0.40277400054392165,
      "grad_norm": 0.37109375,
      "learning_rate": 1.4916164139810343e-05,
      "loss": 0.6364,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2962
    },
    {
      "epoch": 0.4029099809627414,
      "grad_norm": 0.3984375,
      "learning_rate": 1.491202918033331e-05,
      "loss": 0.6522,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2963
    },
    {
      "epoch": 0.40304596138156107,
      "grad_norm": 0.341796875,
      "learning_rate": 1.4907893113624844e-05,
      "loss": 0.5752,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2964
    },
    {
      "epoch": 0.40318194180038075,
      "grad_norm": 0.259765625,
      "learning_rate": 1.4903755940617269e-05,
      "loss": 0.465,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2965
    },
    {
      "epoch": 0.40331792221920043,
      "grad_norm": 0.251953125,
      "learning_rate": 1.4899617662243149e-05,
      "loss": 0.4082,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2966
    },
    {
      "epoch": 0.4034539026380201,
      "grad_norm": 1.015625,
      "learning_rate": 1.4895478279435305e-05,
      "loss": 0.9027,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2967
    },
    {
      "epoch": 0.4035898830568398,
      "grad_norm": 0.61328125,
      "learning_rate": 1.4891337793126804e-05,
      "loss": 0.6916,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2968
    },
    {
      "epoch": 0.40372586347565953,
      "grad_norm": 0.33203125,
      "learning_rate": 1.4887196204250965e-05,
      "loss": 0.7126,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2969
    },
    {
      "epoch": 0.4038618438944792,
      "grad_norm": 0.296875,
      "learning_rate": 1.4883053513741347e-05,
      "loss": 0.5296,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2970
    },
    {
      "epoch": 0.4039978243132989,
      "grad_norm": 0.26953125,
      "learning_rate": 1.4878909722531768e-05,
      "loss": 0.4725,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2971
    },
    {
      "epoch": 0.4041338047321186,
      "grad_norm": 0.4296875,
      "learning_rate": 1.4874764831556285e-05,
      "loss": 0.756,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2972
    },
    {
      "epoch": 0.40426978515093825,
      "grad_norm": 0.314453125,
      "learning_rate": 1.4870618841749214e-05,
      "loss": 0.605,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2973
    },
    {
      "epoch": 0.40440576556975794,
      "grad_norm": 0.37109375,
      "learning_rate": 1.4866471754045106e-05,
      "loss": 0.6493,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2974
    },
    {
      "epoch": 0.4045417459885777,
      "grad_norm": 0.66796875,
      "learning_rate": 1.4862323569378765e-05,
      "loss": 0.84,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2975
    },
    {
      "epoch": 0.40467772640739735,
      "grad_norm": 0.47265625,
      "learning_rate": 1.4858174288685247e-05,
      "loss": 0.6562,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2976
    },
    {
      "epoch": 0.40481370682621703,
      "grad_norm": 0.58984375,
      "learning_rate": 1.4854023912899847e-05,
      "loss": 0.7114,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2977
    },
    {
      "epoch": 0.4049496872450367,
      "grad_norm": 0.4921875,
      "learning_rate": 1.484987244295811e-05,
      "loss": 0.8411,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2978
    },
    {
      "epoch": 0.4050856676638564,
      "grad_norm": 0.32421875,
      "learning_rate": 1.484571987979583e-05,
      "loss": 0.5609,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2979
    },
    {
      "epoch": 0.4052216480826761,
      "grad_norm": 0.447265625,
      "learning_rate": 1.4841566224349047e-05,
      "loss": 0.564,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2980
    },
    {
      "epoch": 0.40535762850149576,
      "grad_norm": 0.2294921875,
      "learning_rate": 1.483741147755404e-05,
      "loss": 0.4705,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2981
    },
    {
      "epoch": 0.4054936089203155,
      "grad_norm": 0.578125,
      "learning_rate": 1.4833255640347346e-05,
      "loss": 0.8089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2982
    },
    {
      "epoch": 0.4056295893391352,
      "grad_norm": 0.41015625,
      "learning_rate": 1.4829098713665739e-05,
      "loss": 0.7199,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2983
    },
    {
      "epoch": 0.40576556975795486,
      "grad_norm": 0.1767578125,
      "learning_rate": 1.482494069844624e-05,
      "loss": 0.3133,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2984
    },
    {
      "epoch": 0.40590155017677454,
      "grad_norm": 0.2490234375,
      "learning_rate": 1.4820781595626116e-05,
      "loss": 0.4285,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2985
    },
    {
      "epoch": 0.4060375305955942,
      "grad_norm": 0.49609375,
      "learning_rate": 1.4816621406142883e-05,
      "loss": 0.8223,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2986
    },
    {
      "epoch": 0.4061735110144139,
      "grad_norm": 0.310546875,
      "learning_rate": 1.4812460130934297e-05,
      "loss": 0.5744,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2987
    },
    {
      "epoch": 0.40630949143323364,
      "grad_norm": 0.48046875,
      "learning_rate": 1.480829777093836e-05,
      "loss": 0.9087,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2988
    },
    {
      "epoch": 0.4064454718520533,
      "grad_norm": 0.388671875,
      "learning_rate": 1.4804134327093318e-05,
      "loss": 0.6715,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2989
    },
    {
      "epoch": 0.406581452270873,
      "grad_norm": 0.5625,
      "learning_rate": 1.4799969800337663e-05,
      "loss": 0.5966,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2990
    },
    {
      "epoch": 0.4067174326896927,
      "grad_norm": 0.494140625,
      "learning_rate": 1.4795804191610131e-05,
      "loss": 0.7578,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2991
    },
    {
      "epoch": 0.40685341310851236,
      "grad_norm": 0.53515625,
      "learning_rate": 1.4791637501849698e-05,
      "loss": 0.5728,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2992
    },
    {
      "epoch": 0.40698939352733204,
      "grad_norm": 0.365234375,
      "learning_rate": 1.4787469731995592e-05,
      "loss": 0.6943,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2993
    },
    {
      "epoch": 0.4071253739461518,
      "grad_norm": 0.298828125,
      "learning_rate": 1.4783300882987275e-05,
      "loss": 0.4992,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2994
    },
    {
      "epoch": 0.40726135436497146,
      "grad_norm": 0.474609375,
      "learning_rate": 1.4779130955764458e-05,
      "loss": 0.5651,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2995
    },
    {
      "epoch": 0.40739733478379114,
      "grad_norm": 0.404296875,
      "learning_rate": 1.4774959951267093e-05,
      "loss": 0.5068,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2996
    },
    {
      "epoch": 0.4075333152026108,
      "grad_norm": 0.3671875,
      "learning_rate": 1.4770787870435377e-05,
      "loss": 0.614,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2997
    },
    {
      "epoch": 0.4076692956214305,
      "grad_norm": 0.275390625,
      "learning_rate": 1.4766614714209748e-05,
      "loss": 0.603,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2998
    },
    {
      "epoch": 0.4078052760402502,
      "grad_norm": 0.515625,
      "learning_rate": 1.4762440483530883e-05,
      "loss": 0.7064,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 2999
    },
    {
      "epoch": 0.40794125645906987,
      "grad_norm": 0.38671875,
      "learning_rate": 1.4758265179339707e-05,
      "loss": 0.5901,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3000
    },
    {
      "epoch": 0.4080772368778896,
      "grad_norm": 0.314453125,
      "learning_rate": 1.4754088802577387e-05,
      "loss": 0.6592,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3001
    },
    {
      "epoch": 0.4082132172967093,
      "grad_norm": 0.380859375,
      "learning_rate": 1.4749911354185327e-05,
      "loss": 0.8311,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3002
    },
    {
      "epoch": 0.40834919771552897,
      "grad_norm": 0.337890625,
      "learning_rate": 1.4745732835105173e-05,
      "loss": 0.5758,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3003
    },
    {
      "epoch": 0.40848517813434865,
      "grad_norm": 0.53125,
      "learning_rate": 1.4741553246278817e-05,
      "loss": 0.4665,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3004
    },
    {
      "epoch": 0.40862115855316833,
      "grad_norm": 0.458984375,
      "learning_rate": 1.4737372588648388e-05,
      "loss": 0.7134,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3005
    },
    {
      "epoch": 0.408757138971988,
      "grad_norm": 0.478515625,
      "learning_rate": 1.473319086315626e-05,
      "loss": 0.6997,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3006
    },
    {
      "epoch": 0.40889311939080775,
      "grad_norm": 0.53515625,
      "learning_rate": 1.4729008070745038e-05,
      "loss": 0.5828,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3007
    },
    {
      "epoch": 0.4090290998096274,
      "grad_norm": 0.3671875,
      "learning_rate": 1.4724824212357583e-05,
      "loss": 0.5422,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3008
    },
    {
      "epoch": 0.4091650802284471,
      "grad_norm": 0.27734375,
      "learning_rate": 1.4720639288936985e-05,
      "loss": 0.467,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3009
    },
    {
      "epoch": 0.4093010606472668,
      "grad_norm": 0.419921875,
      "learning_rate": 1.4716453301426576e-05,
      "loss": 0.6095,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3010
    },
    {
      "epoch": 0.40943704106608647,
      "grad_norm": 0.44140625,
      "learning_rate": 1.4712266250769932e-05,
      "loss": 0.6929,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3011
    },
    {
      "epoch": 0.40957302148490615,
      "grad_norm": 0.373046875,
      "learning_rate": 1.4708078137910857e-05,
      "loss": 0.6146,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3012
    },
    {
      "epoch": 0.4097090019037259,
      "grad_norm": 0.37890625,
      "learning_rate": 1.4703888963793414e-05,
      "loss": 0.6309,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3013
    },
    {
      "epoch": 0.40984498232254557,
      "grad_norm": 0.337890625,
      "learning_rate": 1.4699698729361884e-05,
      "loss": 0.6637,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3014
    },
    {
      "epoch": 0.40998096274136525,
      "grad_norm": 0.435546875,
      "learning_rate": 1.4695507435560804e-05,
      "loss": 0.7677,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3015
    },
    {
      "epoch": 0.41011694316018493,
      "grad_norm": 0.314453125,
      "learning_rate": 1.4691315083334939e-05,
      "loss": 0.4762,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3016
    },
    {
      "epoch": 0.4102529235790046,
      "grad_norm": 0.353515625,
      "learning_rate": 1.4687121673629299e-05,
      "loss": 0.695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3017
    },
    {
      "epoch": 0.4103889039978243,
      "grad_norm": 0.44140625,
      "learning_rate": 1.4682927207389126e-05,
      "loss": 0.7949,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3018
    },
    {
      "epoch": 0.410524884416644,
      "grad_norm": 0.462890625,
      "learning_rate": 1.4678731685559911e-05,
      "loss": 0.6335,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3019
    },
    {
      "epoch": 0.4106608648354637,
      "grad_norm": 0.6328125,
      "learning_rate": 1.4674535109087366e-05,
      "loss": 0.7618,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3020
    },
    {
      "epoch": 0.4107968452542834,
      "grad_norm": 0.50390625,
      "learning_rate": 1.4670337478917462e-05,
      "loss": 0.4572,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3021
    },
    {
      "epoch": 0.4109328256731031,
      "grad_norm": 0.34375,
      "learning_rate": 1.4666138795996387e-05,
      "loss": 0.6328,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3022
    },
    {
      "epoch": 0.41106880609192276,
      "grad_norm": 0.2412109375,
      "learning_rate": 1.466193906127058e-05,
      "loss": 0.3234,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3023
    },
    {
      "epoch": 0.41120478651074244,
      "grad_norm": 0.32421875,
      "learning_rate": 1.4657738275686711e-05,
      "loss": 0.6079,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3024
    },
    {
      "epoch": 0.4113407669295621,
      "grad_norm": 0.4921875,
      "learning_rate": 1.4653536440191687e-05,
      "loss": 0.7049,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3025
    },
    {
      "epoch": 0.41147674734838185,
      "grad_norm": 0.314453125,
      "learning_rate": 1.4649333555732656e-05,
      "loss": 0.5518,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3026
    },
    {
      "epoch": 0.41161272776720154,
      "grad_norm": 0.72265625,
      "learning_rate": 1.4645129623257e-05,
      "loss": 0.7435,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3027
    },
    {
      "epoch": 0.4117487081860212,
      "grad_norm": 0.494140625,
      "learning_rate": 1.4640924643712336e-05,
      "loss": 0.8302,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3028
    },
    {
      "epoch": 0.4118846886048409,
      "grad_norm": 0.56640625,
      "learning_rate": 1.4636718618046512e-05,
      "loss": 0.6335,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3029
    },
    {
      "epoch": 0.4120206690236606,
      "grad_norm": 0.3359375,
      "learning_rate": 1.4632511547207629e-05,
      "loss": 0.6012,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3030
    },
    {
      "epoch": 0.41215664944248026,
      "grad_norm": 0.466796875,
      "learning_rate": 1.4628303432144002e-05,
      "loss": 0.8123,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3031
    },
    {
      "epoch": 0.4122926298613,
      "grad_norm": 0.48828125,
      "learning_rate": 1.4624094273804196e-05,
      "loss": 0.6232,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3032
    },
    {
      "epoch": 0.4124286102801197,
      "grad_norm": 0.26953125,
      "learning_rate": 1.4619884073137006e-05,
      "loss": 0.4746,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3033
    },
    {
      "epoch": 0.41256459069893936,
      "grad_norm": 0.546875,
      "learning_rate": 1.4615672831091463e-05,
      "loss": 0.7739,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3034
    },
    {
      "epoch": 0.41270057111775904,
      "grad_norm": 0.46484375,
      "learning_rate": 1.461146054861683e-05,
      "loss": 0.7041,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3035
    },
    {
      "epoch": 0.4128365515365787,
      "grad_norm": 0.443359375,
      "learning_rate": 1.4607247226662613e-05,
      "loss": 0.6941,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3036
    },
    {
      "epoch": 0.4129725319553984,
      "grad_norm": 0.470703125,
      "learning_rate": 1.460303286617854e-05,
      "loss": 0.7713,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3037
    },
    {
      "epoch": 0.41310851237421814,
      "grad_norm": 0.57421875,
      "learning_rate": 1.4598817468114583e-05,
      "loss": 0.626,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3038
    },
    {
      "epoch": 0.4132444927930378,
      "grad_norm": 0.3984375,
      "learning_rate": 1.4594601033420944e-05,
      "loss": 0.7622,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3039
    },
    {
      "epoch": 0.4133804732118575,
      "grad_norm": 0.275390625,
      "learning_rate": 1.4590383563048055e-05,
      "loss": 0.4816,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3040
    },
    {
      "epoch": 0.4135164536306772,
      "grad_norm": 0.271484375,
      "learning_rate": 1.4586165057946591e-05,
      "loss": 0.3843,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3041
    },
    {
      "epoch": 0.41365243404949686,
      "grad_norm": 0.2890625,
      "learning_rate": 1.4581945519067452e-05,
      "loss": 0.3758,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3042
    },
    {
      "epoch": 0.41378841446831655,
      "grad_norm": 0.515625,
      "learning_rate": 1.4577724947361774e-05,
      "loss": 0.583,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3043
    },
    {
      "epoch": 0.4139243948871362,
      "grad_norm": 0.5234375,
      "learning_rate": 1.4573503343780924e-05,
      "loss": 0.8211,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3044
    },
    {
      "epoch": 0.41406037530595596,
      "grad_norm": 0.23046875,
      "learning_rate": 1.4569280709276506e-05,
      "loss": 0.4341,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3045
    },
    {
      "epoch": 0.41419635572477564,
      "grad_norm": 0.33984375,
      "learning_rate": 1.456505704480035e-05,
      "loss": 0.6886,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3046
    },
    {
      "epoch": 0.4143323361435953,
      "grad_norm": 0.330078125,
      "learning_rate": 1.4560832351304523e-05,
      "loss": 0.3815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3047
    },
    {
      "epoch": 0.414468316562415,
      "grad_norm": 0.232421875,
      "learning_rate": 1.4556606629741326e-05,
      "loss": 0.4043,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3048
    },
    {
      "epoch": 0.4146042969812347,
      "grad_norm": 0.37890625,
      "learning_rate": 1.4552379881063283e-05,
      "loss": 0.7518,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3049
    },
    {
      "epoch": 0.41474027740005437,
      "grad_norm": 0.291015625,
      "learning_rate": 1.4548152106223157e-05,
      "loss": 0.5369,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3050
    },
    {
      "epoch": 0.4148762578188741,
      "grad_norm": 0.5078125,
      "learning_rate": 1.454392330617394e-05,
      "loss": 0.5814,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3051
    },
    {
      "epoch": 0.4150122382376938,
      "grad_norm": 0.44140625,
      "learning_rate": 1.4539693481868858e-05,
      "loss": 0.7544,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3052
    },
    {
      "epoch": 0.41514821865651347,
      "grad_norm": 0.35546875,
      "learning_rate": 1.4535462634261359e-05,
      "loss": 0.6536,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3053
    },
    {
      "epoch": 0.41528419907533315,
      "grad_norm": 0.267578125,
      "learning_rate": 1.4531230764305133e-05,
      "loss": 0.6021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3054
    },
    {
      "epoch": 0.41542017949415283,
      "grad_norm": 0.302734375,
      "learning_rate": 1.4526997872954094e-05,
      "loss": 0.5824,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3055
    },
    {
      "epoch": 0.4155561599129725,
      "grad_norm": 0.828125,
      "learning_rate": 1.4522763961162385e-05,
      "loss": 0.7049,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3056
    },
    {
      "epoch": 0.41569214033179225,
      "grad_norm": 0.578125,
      "learning_rate": 1.4518529029884386e-05,
      "loss": 0.6701,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3057
    },
    {
      "epoch": 0.41582812075061193,
      "grad_norm": 0.28125,
      "learning_rate": 1.4514293080074697e-05,
      "loss": 0.5763,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3058
    },
    {
      "epoch": 0.4159641011694316,
      "grad_norm": 0.3359375,
      "learning_rate": 1.4510056112688159e-05,
      "loss": 0.591,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3059
    },
    {
      "epoch": 0.4161000815882513,
      "grad_norm": 0.423828125,
      "learning_rate": 1.450581812867983e-05,
      "loss": 0.8162,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3060
    },
    {
      "epoch": 0.416236062007071,
      "grad_norm": 0.70703125,
      "learning_rate": 1.4501579129005008e-05,
      "loss": 0.5923,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3061
    },
    {
      "epoch": 0.41637204242589065,
      "grad_norm": 0.34765625,
      "learning_rate": 1.4497339114619214e-05,
      "loss": 0.5651,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3062
    },
    {
      "epoch": 0.41650802284471033,
      "grad_norm": 0.234375,
      "learning_rate": 1.4493098086478196e-05,
      "loss": 0.4471,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3063
    },
    {
      "epoch": 0.41664400326353007,
      "grad_norm": 0.4140625,
      "learning_rate": 1.4488856045537937e-05,
      "loss": 0.6746,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3064
    },
    {
      "epoch": 0.41677998368234975,
      "grad_norm": 0.455078125,
      "learning_rate": 1.4484612992754647e-05,
      "loss": 0.6989,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3065
    },
    {
      "epoch": 0.41691596410116943,
      "grad_norm": 0.39453125,
      "learning_rate": 1.4480368929084758e-05,
      "loss": 0.6852,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3066
    },
    {
      "epoch": 0.4170519445199891,
      "grad_norm": 0.421875,
      "learning_rate": 1.4476123855484931e-05,
      "loss": 0.71,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3067
    },
    {
      "epoch": 0.4171879249388088,
      "grad_norm": 0.484375,
      "learning_rate": 1.4471877772912067e-05,
      "loss": 0.707,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3068
    },
    {
      "epoch": 0.4173239053576285,
      "grad_norm": 0.396484375,
      "learning_rate": 1.4467630682323276e-05,
      "loss": 0.7103,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3069
    },
    {
      "epoch": 0.4174598857764482,
      "grad_norm": 0.466796875,
      "learning_rate": 1.4463382584675908e-05,
      "loss": 0.6206,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3070
    },
    {
      "epoch": 0.4175958661952679,
      "grad_norm": 0.2734375,
      "learning_rate": 1.4459133480927539e-05,
      "loss": 0.4935,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3071
    },
    {
      "epoch": 0.4177318466140876,
      "grad_norm": 0.380859375,
      "learning_rate": 1.4454883372035964e-05,
      "loss": 0.7091,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3072
    },
    {
      "epoch": 0.41786782703290726,
      "grad_norm": 0.404296875,
      "learning_rate": 1.445063225895921e-05,
      "loss": 0.7796,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3073
    },
    {
      "epoch": 0.41800380745172694,
      "grad_norm": 0.33203125,
      "learning_rate": 1.4446380142655532e-05,
      "loss": 0.6226,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3074
    },
    {
      "epoch": 0.4181397878705466,
      "grad_norm": 0.412109375,
      "learning_rate": 1.4442127024083407e-05,
      "loss": 0.6877,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3075
    },
    {
      "epoch": 0.41827576828936636,
      "grad_norm": 0.400390625,
      "learning_rate": 1.4437872904201542e-05,
      "loss": 0.6973,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3076
    },
    {
      "epoch": 0.41841174870818604,
      "grad_norm": 0.341796875,
      "learning_rate": 1.4433617783968868e-05,
      "loss": 0.6323,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3077
    },
    {
      "epoch": 0.4185477291270057,
      "grad_norm": 0.2216796875,
      "learning_rate": 1.4429361664344538e-05,
      "loss": 0.4355,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3078
    },
    {
      "epoch": 0.4186837095458254,
      "grad_norm": 0.4296875,
      "learning_rate": 1.4425104546287934e-05,
      "loss": 0.7809,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3079
    },
    {
      "epoch": 0.4188196899646451,
      "grad_norm": 0.494140625,
      "learning_rate": 1.4420846430758666e-05,
      "loss": 0.6714,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3080
    },
    {
      "epoch": 0.41895567038346476,
      "grad_norm": 0.490234375,
      "learning_rate": 1.4416587318716561e-05,
      "loss": 0.7357,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3081
    },
    {
      "epoch": 0.41909165080228444,
      "grad_norm": 0.3046875,
      "learning_rate": 1.4412327211121678e-05,
      "loss": 0.5457,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3082
    },
    {
      "epoch": 0.4192276312211042,
      "grad_norm": 0.3671875,
      "learning_rate": 1.4408066108934296e-05,
      "loss": 0.6131,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3083
    },
    {
      "epoch": 0.41936361163992386,
      "grad_norm": 1.34375,
      "learning_rate": 1.4403804013114921e-05,
      "loss": 0.8003,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3084
    },
    {
      "epoch": 0.41949959205874354,
      "grad_norm": 0.63671875,
      "learning_rate": 1.4399540924624278e-05,
      "loss": 0.724,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3085
    },
    {
      "epoch": 0.4196355724775632,
      "grad_norm": 0.302734375,
      "learning_rate": 1.4395276844423322e-05,
      "loss": 0.6592,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3086
    },
    {
      "epoch": 0.4197715528963829,
      "grad_norm": 0.458984375,
      "learning_rate": 1.439101177347323e-05,
      "loss": 0.8862,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3087
    },
    {
      "epoch": 0.4199075333152026,
      "grad_norm": 0.36328125,
      "learning_rate": 1.4386745712735398e-05,
      "loss": 0.6701,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3088
    },
    {
      "epoch": 0.4200435137340223,
      "grad_norm": 0.515625,
      "learning_rate": 1.4382478663171449e-05,
      "loss": 0.8861,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3089
    },
    {
      "epoch": 0.420179494152842,
      "grad_norm": 0.40234375,
      "learning_rate": 1.4378210625743228e-05,
      "loss": 0.6484,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3090
    },
    {
      "epoch": 0.4203154745716617,
      "grad_norm": 0.349609375,
      "learning_rate": 1.4373941601412807e-05,
      "loss": 0.6969,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3091
    },
    {
      "epoch": 0.42045145499048137,
      "grad_norm": 0.69921875,
      "learning_rate": 1.4369671591142469e-05,
      "loss": 0.8148,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3092
    },
    {
      "epoch": 0.42058743540930105,
      "grad_norm": 0.2060546875,
      "learning_rate": 1.4365400595894732e-05,
      "loss": 0.3413,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3093
    },
    {
      "epoch": 0.42072341582812073,
      "grad_norm": 0.66796875,
      "learning_rate": 1.4361128616632329e-05,
      "loss": 0.8372,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3094
    },
    {
      "epoch": 0.42085939624694046,
      "grad_norm": 0.26953125,
      "learning_rate": 1.4356855654318215e-05,
      "loss": 0.6174,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3095
    },
    {
      "epoch": 0.42099537666576015,
      "grad_norm": 0.63671875,
      "learning_rate": 1.4352581709915572e-05,
      "loss": 0.646,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3096
    },
    {
      "epoch": 0.4211313570845798,
      "grad_norm": 0.44140625,
      "learning_rate": 1.4348306784387796e-05,
      "loss": 0.6401,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3097
    },
    {
      "epoch": 0.4212673375033995,
      "grad_norm": 0.39453125,
      "learning_rate": 1.4344030878698506e-05,
      "loss": 0.7096,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3098
    },
    {
      "epoch": 0.4214033179222192,
      "grad_norm": 0.84375,
      "learning_rate": 1.4339753993811545e-05,
      "loss": 0.7539,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3099
    },
    {
      "epoch": 0.42153929834103887,
      "grad_norm": 0.41015625,
      "learning_rate": 1.4335476130690978e-05,
      "loss": 0.6834,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3100
    },
    {
      "epoch": 0.4216752787598586,
      "grad_norm": 0.59375,
      "learning_rate": 1.4331197290301084e-05,
      "loss": 0.8639,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3101
    },
    {
      "epoch": 0.4218112591786783,
      "grad_norm": 0.75390625,
      "learning_rate": 1.4326917473606368e-05,
      "loss": 0.765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3102
    },
    {
      "epoch": 0.42194723959749797,
      "grad_norm": 0.384765625,
      "learning_rate": 1.4322636681571552e-05,
      "loss": 0.6766,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3103
    },
    {
      "epoch": 0.42208322001631765,
      "grad_norm": 0.33984375,
      "learning_rate": 1.4318354915161578e-05,
      "loss": 0.6958,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3104
    },
    {
      "epoch": 0.42221920043513733,
      "grad_norm": 0.390625,
      "learning_rate": 1.4314072175341611e-05,
      "loss": 0.7438,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3105
    },
    {
      "epoch": 0.422355180853957,
      "grad_norm": 0.314453125,
      "learning_rate": 1.4309788463077033e-05,
      "loss": 0.5329,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3106
    },
    {
      "epoch": 0.4224911612727767,
      "grad_norm": 0.357421875,
      "learning_rate": 1.4305503779333441e-05,
      "loss": 0.7113,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3107
    },
    {
      "epoch": 0.42262714169159643,
      "grad_norm": 0.388671875,
      "learning_rate": 1.4301218125076656e-05,
      "loss": 0.6338,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3108
    },
    {
      "epoch": 0.4227631221104161,
      "grad_norm": 0.34375,
      "learning_rate": 1.4296931501272721e-05,
      "loss": 0.6549,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3109
    },
    {
      "epoch": 0.4228991025292358,
      "grad_norm": 0.4921875,
      "learning_rate": 1.4292643908887893e-05,
      "loss": 0.6248,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3110
    },
    {
      "epoch": 0.4230350829480555,
      "grad_norm": 0.39453125,
      "learning_rate": 1.4288355348888643e-05,
      "loss": 0.6214,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3111
    },
    {
      "epoch": 0.42317106336687516,
      "grad_norm": 0.349609375,
      "learning_rate": 1.4284065822241669e-05,
      "loss": 0.4924,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3112
    },
    {
      "epoch": 0.42330704378569484,
      "grad_norm": 0.423828125,
      "learning_rate": 1.4279775329913878e-05,
      "loss": 0.6886,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3113
    },
    {
      "epoch": 0.4234430242045146,
      "grad_norm": 0.267578125,
      "learning_rate": 1.4275483872872407e-05,
      "loss": 0.5378,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3114
    },
    {
      "epoch": 0.42357900462333425,
      "grad_norm": 0.470703125,
      "learning_rate": 1.4271191452084598e-05,
      "loss": 0.6966,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3115
    },
    {
      "epoch": 0.42371498504215394,
      "grad_norm": 0.9375,
      "learning_rate": 1.4266898068518013e-05,
      "loss": 0.9907,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3116
    },
    {
      "epoch": 0.4238509654609736,
      "grad_norm": 0.4609375,
      "learning_rate": 1.4262603723140438e-05,
      "loss": 0.4619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3117
    },
    {
      "epoch": 0.4239869458797933,
      "grad_norm": 0.41015625,
      "learning_rate": 1.4258308416919868e-05,
      "loss": 0.5962,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3118
    },
    {
      "epoch": 0.424122926298613,
      "grad_norm": 0.59765625,
      "learning_rate": 1.4254012150824522e-05,
      "loss": 0.492,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3119
    },
    {
      "epoch": 0.4242589067174327,
      "grad_norm": 1.390625,
      "learning_rate": 1.4249714925822824e-05,
      "loss": 0.903,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3120
    },
    {
      "epoch": 0.4243948871362524,
      "grad_norm": 0.34765625,
      "learning_rate": 1.4245416742883422e-05,
      "loss": 0.5871,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3121
    },
    {
      "epoch": 0.4245308675550721,
      "grad_norm": 0.4609375,
      "learning_rate": 1.4241117602975186e-05,
      "loss": 0.4814,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3122
    },
    {
      "epoch": 0.42466684797389176,
      "grad_norm": 0.39453125,
      "learning_rate": 1.4236817507067189e-05,
      "loss": 0.7089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3123
    },
    {
      "epoch": 0.42480282839271144,
      "grad_norm": 0.62890625,
      "learning_rate": 1.4232516456128725e-05,
      "loss": 0.7145,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3124
    },
    {
      "epoch": 0.4249388088115311,
      "grad_norm": 0.59765625,
      "learning_rate": 1.4228214451129306e-05,
      "loss": 0.9424,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3125
    },
    {
      "epoch": 0.4250747892303508,
      "grad_norm": 0.39453125,
      "learning_rate": 1.422391149303865e-05,
      "loss": 0.8651,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3126
    },
    {
      "epoch": 0.42521076964917054,
      "grad_norm": 0.30859375,
      "learning_rate": 1.421960758282671e-05,
      "loss": 0.6387,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3127
    },
    {
      "epoch": 0.4253467500679902,
      "grad_norm": 0.486328125,
      "learning_rate": 1.4215302721463624e-05,
      "loss": 0.5686,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3128
    },
    {
      "epoch": 0.4254827304868099,
      "grad_norm": 0.40625,
      "learning_rate": 1.4210996909919769e-05,
      "loss": 0.7134,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3129
    },
    {
      "epoch": 0.4256187109056296,
      "grad_norm": 0.54296875,
      "learning_rate": 1.4206690149165731e-05,
      "loss": 0.8548,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3130
    },
    {
      "epoch": 0.42575469132444926,
      "grad_norm": 0.55859375,
      "learning_rate": 1.42023824401723e-05,
      "loss": 0.6069,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3131
    },
    {
      "epoch": 0.42589067174326894,
      "grad_norm": 0.55078125,
      "learning_rate": 1.4198073783910488e-05,
      "loss": 0.5871,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3132
    },
    {
      "epoch": 0.4260266521620887,
      "grad_norm": 0.51171875,
      "learning_rate": 1.4193764181351519e-05,
      "loss": 0.5169,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3133
    },
    {
      "epoch": 0.42616263258090836,
      "grad_norm": 0.5234375,
      "learning_rate": 1.418945363346683e-05,
      "loss": 0.7511,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3134
    },
    {
      "epoch": 0.42629861299972804,
      "grad_norm": 0.29296875,
      "learning_rate": 1.4185142141228072e-05,
      "loss": 0.5177,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3135
    },
    {
      "epoch": 0.4264345934185477,
      "grad_norm": 0.37109375,
      "learning_rate": 1.418082970560711e-05,
      "loss": 0.5568,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3136
    },
    {
      "epoch": 0.4265705738373674,
      "grad_norm": 0.37890625,
      "learning_rate": 1.4176516327576014e-05,
      "loss": 0.6214,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3137
    },
    {
      "epoch": 0.4267065542561871,
      "grad_norm": 0.396484375,
      "learning_rate": 1.4172202008107076e-05,
      "loss": 0.5851,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3138
    },
    {
      "epoch": 0.4268425346750068,
      "grad_norm": 0.380859375,
      "learning_rate": 1.4167886748172797e-05,
      "loss": 0.6571,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3139
    },
    {
      "epoch": 0.4269785150938265,
      "grad_norm": 0.478515625,
      "learning_rate": 1.4163570548745886e-05,
      "loss": 0.6455,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3140
    },
    {
      "epoch": 0.4271144955126462,
      "grad_norm": 0.5390625,
      "learning_rate": 1.4159253410799272e-05,
      "loss": 0.5622,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3141
    },
    {
      "epoch": 0.42725047593146587,
      "grad_norm": 0.458984375,
      "learning_rate": 1.4154935335306084e-05,
      "loss": 0.8169,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3142
    },
    {
      "epoch": 0.42738645635028555,
      "grad_norm": 0.2373046875,
      "learning_rate": 1.4150616323239677e-05,
      "loss": 0.4801,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3143
    },
    {
      "epoch": 0.42752243676910523,
      "grad_norm": 0.326171875,
      "learning_rate": 1.41462963755736e-05,
      "loss": 0.5978,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3144
    },
    {
      "epoch": 0.42765841718792497,
      "grad_norm": 0.44140625,
      "learning_rate": 1.414197549328163e-05,
      "loss": 0.5617,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3145
    },
    {
      "epoch": 0.42779439760674465,
      "grad_norm": 0.294921875,
      "learning_rate": 1.4137653677337743e-05,
      "loss": 0.5027,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3146
    },
    {
      "epoch": 0.42793037802556433,
      "grad_norm": 0.412109375,
      "learning_rate": 1.4133330928716132e-05,
      "loss": 0.7606,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3147
    },
    {
      "epoch": 0.428066358444384,
      "grad_norm": 0.41796875,
      "learning_rate": 1.4129007248391193e-05,
      "loss": 0.5588,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3148
    },
    {
      "epoch": 0.4282023388632037,
      "grad_norm": 0.322265625,
      "learning_rate": 1.4124682637337538e-05,
      "loss": 0.4953,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3149
    },
    {
      "epoch": 0.42833831928202337,
      "grad_norm": 0.357421875,
      "learning_rate": 1.412035709652999e-05,
      "loss": 0.528,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3150
    },
    {
      "epoch": 0.42847429970084305,
      "grad_norm": 0.55078125,
      "learning_rate": 1.4116030626943576e-05,
      "loss": 0.6704,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3151
    },
    {
      "epoch": 0.4286102801196628,
      "grad_norm": 0.291015625,
      "learning_rate": 1.4111703229553536e-05,
      "loss": 0.6686,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3152
    },
    {
      "epoch": 0.42874626053848247,
      "grad_norm": 0.455078125,
      "learning_rate": 1.4107374905335316e-05,
      "loss": 0.5565,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3153
    },
    {
      "epoch": 0.42888224095730215,
      "grad_norm": 0.4140625,
      "learning_rate": 1.4103045655264576e-05,
      "loss": 0.7463,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3154
    },
    {
      "epoch": 0.42901822137612183,
      "grad_norm": 0.37890625,
      "learning_rate": 1.4098715480317182e-05,
      "loss": 0.6683,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3155
    },
    {
      "epoch": 0.4291542017949415,
      "grad_norm": 0.310546875,
      "learning_rate": 1.4094384381469209e-05,
      "loss": 0.6221,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3156
    },
    {
      "epoch": 0.4292901822137612,
      "grad_norm": 0.36328125,
      "learning_rate": 1.4090052359696935e-05,
      "loss": 0.5749,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3157
    },
    {
      "epoch": 0.42942616263258093,
      "grad_norm": 0.671875,
      "learning_rate": 1.4085719415976852e-05,
      "loss": 0.6336,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3158
    },
    {
      "epoch": 0.4295621430514006,
      "grad_norm": 0.40625,
      "learning_rate": 1.4081385551285664e-05,
      "loss": 0.5244,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3159
    },
    {
      "epoch": 0.4296981234702203,
      "grad_norm": 0.4921875,
      "learning_rate": 1.4077050766600273e-05,
      "loss": 0.6128,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3160
    },
    {
      "epoch": 0.42983410388904,
      "grad_norm": 0.4140625,
      "learning_rate": 1.4072715062897793e-05,
      "loss": 0.6994,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3161
    },
    {
      "epoch": 0.42997008430785966,
      "grad_norm": 0.4296875,
      "learning_rate": 1.4068378441155544e-05,
      "loss": 0.6561,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3162
    },
    {
      "epoch": 0.43010606472667934,
      "grad_norm": 0.470703125,
      "learning_rate": 1.4064040902351055e-05,
      "loss": 0.721,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3163
    },
    {
      "epoch": 0.4302420451454991,
      "grad_norm": 0.5234375,
      "learning_rate": 1.4059702447462059e-05,
      "loss": 0.5426,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3164
    },
    {
      "epoch": 0.43037802556431876,
      "grad_norm": 0.244140625,
      "learning_rate": 1.4055363077466497e-05,
      "loss": 0.4284,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3165
    },
    {
      "epoch": 0.43051400598313844,
      "grad_norm": 0.359375,
      "learning_rate": 1.4051022793342515e-05,
      "loss": 0.6978,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3166
    },
    {
      "epoch": 0.4306499864019581,
      "grad_norm": 0.296875,
      "learning_rate": 1.4046681596068468e-05,
      "loss": 0.5936,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3167
    },
    {
      "epoch": 0.4307859668207778,
      "grad_norm": 0.443359375,
      "learning_rate": 1.4042339486622913e-05,
      "loss": 0.597,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3168
    },
    {
      "epoch": 0.4309219472395975,
      "grad_norm": 0.44921875,
      "learning_rate": 1.4037996465984618e-05,
      "loss": 0.6214,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3169
    },
    {
      "epoch": 0.43105792765841716,
      "grad_norm": 0.455078125,
      "learning_rate": 1.4033652535132549e-05,
      "loss": 0.7832,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3170
    },
    {
      "epoch": 0.4311939080772369,
      "grad_norm": 0.357421875,
      "learning_rate": 1.4029307695045884e-05,
      "loss": 0.6999,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3171
    },
    {
      "epoch": 0.4313298884960566,
      "grad_norm": 0.484375,
      "learning_rate": 1.4024961946704e-05,
      "loss": 0.6995,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3172
    },
    {
      "epoch": 0.43146586891487626,
      "grad_norm": 0.4921875,
      "learning_rate": 1.4020615291086488e-05,
      "loss": 0.7965,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3173
    },
    {
      "epoch": 0.43160184933369594,
      "grad_norm": 0.328125,
      "learning_rate": 1.401626772917313e-05,
      "loss": 0.5656,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3174
    },
    {
      "epoch": 0.4317378297525156,
      "grad_norm": 0.275390625,
      "learning_rate": 1.4011919261943923e-05,
      "loss": 0.5827,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3175
    },
    {
      "epoch": 0.4318738101713353,
      "grad_norm": 0.349609375,
      "learning_rate": 1.4007569890379062e-05,
      "loss": 0.6029,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3176
    },
    {
      "epoch": 0.43200979059015504,
      "grad_norm": 1.1640625,
      "learning_rate": 1.4003219615458954e-05,
      "loss": 0.7808,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3177
    },
    {
      "epoch": 0.4321457710089747,
      "grad_norm": 0.6328125,
      "learning_rate": 1.3998868438164199e-05,
      "loss": 0.7796,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3178
    },
    {
      "epoch": 0.4322817514277944,
      "grad_norm": 0.421875,
      "learning_rate": 1.399451635947561e-05,
      "loss": 0.6943,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3179
    },
    {
      "epoch": 0.4324177318466141,
      "grad_norm": 0.443359375,
      "learning_rate": 1.3990163380374195e-05,
      "loss": 0.6929,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3180
    },
    {
      "epoch": 0.43255371226543377,
      "grad_norm": 2.9375,
      "learning_rate": 1.398580950184117e-05,
      "loss": 0.7948,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3181
    },
    {
      "epoch": 0.43268969268425345,
      "grad_norm": 0.26171875,
      "learning_rate": 1.3981454724857952e-05,
      "loss": 0.4744,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3182
    },
    {
      "epoch": 0.4328256731030732,
      "grad_norm": 0.279296875,
      "learning_rate": 1.3977099050406163e-05,
      "loss": 0.4585,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3183
    },
    {
      "epoch": 0.43296165352189286,
      "grad_norm": 0.470703125,
      "learning_rate": 1.3972742479467622e-05,
      "loss": 0.481,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3184
    },
    {
      "epoch": 0.43309763394071255,
      "grad_norm": 0.396484375,
      "learning_rate": 1.3968385013024355e-05,
      "loss": 0.6242,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3185
    },
    {
      "epoch": 0.4332336143595322,
      "grad_norm": 0.390625,
      "learning_rate": 1.3964026652058589e-05,
      "loss": 0.6969,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3186
    },
    {
      "epoch": 0.4333695947783519,
      "grad_norm": 0.333984375,
      "learning_rate": 1.395966739755275e-05,
      "loss": 0.6379,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3187
    },
    {
      "epoch": 0.4335055751971716,
      "grad_norm": 0.40234375,
      "learning_rate": 1.3955307250489466e-05,
      "loss": 0.7524,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3188
    },
    {
      "epoch": 0.43364155561599127,
      "grad_norm": 0.353515625,
      "learning_rate": 1.3950946211851574e-05,
      "loss": 0.6268,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3189
    },
    {
      "epoch": 0.433777536034811,
      "grad_norm": 0.3125,
      "learning_rate": 1.3946584282622102e-05,
      "loss": 0.686,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3190
    },
    {
      "epoch": 0.4339135164536307,
      "grad_norm": 0.7109375,
      "learning_rate": 1.394222146378428e-05,
      "loss": 0.9004,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3191
    },
    {
      "epoch": 0.43404949687245037,
      "grad_norm": 0.6953125,
      "learning_rate": 1.3937857756321544e-05,
      "loss": 0.6538,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3192
    },
    {
      "epoch": 0.43418547729127005,
      "grad_norm": 0.349609375,
      "learning_rate": 1.3933493161217521e-05,
      "loss": 0.6481,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3193
    },
    {
      "epoch": 0.43432145771008973,
      "grad_norm": 1.5,
      "learning_rate": 1.3929127679456051e-05,
      "loss": 0.7821,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3194
    },
    {
      "epoch": 0.4344574381289094,
      "grad_norm": 0.458984375,
      "learning_rate": 1.3924761312021165e-05,
      "loss": 0.9728,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3195
    },
    {
      "epoch": 0.43459341854772915,
      "grad_norm": 0.45703125,
      "learning_rate": 1.3920394059897093e-05,
      "loss": 0.6387,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3196
    },
    {
      "epoch": 0.43472939896654883,
      "grad_norm": 0.42578125,
      "learning_rate": 1.391602592406827e-05,
      "loss": 0.7362,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3197
    },
    {
      "epoch": 0.4348653793853685,
      "grad_norm": 0.515625,
      "learning_rate": 1.3911656905519326e-05,
      "loss": 0.7638,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3198
    },
    {
      "epoch": 0.4350013598041882,
      "grad_norm": 0.353515625,
      "learning_rate": 1.3907287005235088e-05,
      "loss": 0.6914,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3199
    },
    {
      "epoch": 0.4351373402230079,
      "grad_norm": 0.318359375,
      "learning_rate": 1.3902916224200593e-05,
      "loss": 0.4989,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3200
    },
    {
      "epoch": 0.43527332064182755,
      "grad_norm": 0.35546875,
      "learning_rate": 1.3898544563401062e-05,
      "loss": 0.7401,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3201
    },
    {
      "epoch": 0.4354093010606473,
      "grad_norm": 0.486328125,
      "learning_rate": 1.389417202382192e-05,
      "loss": 0.6637,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3202
    },
    {
      "epoch": 0.435545281479467,
      "grad_norm": 0.240234375,
      "learning_rate": 1.3889798606448797e-05,
      "loss": 0.4222,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3203
    },
    {
      "epoch": 0.43568126189828665,
      "grad_norm": 0.41796875,
      "learning_rate": 1.388542431226751e-05,
      "loss": 0.7694,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3204
    },
    {
      "epoch": 0.43581724231710633,
      "grad_norm": 0.34375,
      "learning_rate": 1.3881049142264078e-05,
      "loss": 0.6102,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3205
    },
    {
      "epoch": 0.435953222735926,
      "grad_norm": 0.416015625,
      "learning_rate": 1.387667309742472e-05,
      "loss": 0.6064,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3206
    },
    {
      "epoch": 0.4360892031547457,
      "grad_norm": 0.318359375,
      "learning_rate": 1.3872296178735851e-05,
      "loss": 0.5199,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3207
    },
    {
      "epoch": 0.43622518357356543,
      "grad_norm": 0.337890625,
      "learning_rate": 1.3867918387184079e-05,
      "loss": 0.6381,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3208
    },
    {
      "epoch": 0.4363611639923851,
      "grad_norm": 0.5625,
      "learning_rate": 1.3863539723756211e-05,
      "loss": 0.4092,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3209
    },
    {
      "epoch": 0.4364971444112048,
      "grad_norm": 0.72265625,
      "learning_rate": 1.3859160189439256e-05,
      "loss": 0.437,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3210
    },
    {
      "epoch": 0.4366331248300245,
      "grad_norm": 0.2373046875,
      "learning_rate": 1.3854779785220412e-05,
      "loss": 0.4232,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3211
    },
    {
      "epoch": 0.43676910524884416,
      "grad_norm": 0.36328125,
      "learning_rate": 1.3850398512087075e-05,
      "loss": 0.7031,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3212
    },
    {
      "epoch": 0.43690508566766384,
      "grad_norm": 0.421875,
      "learning_rate": 1.3846016371026838e-05,
      "loss": 0.5482,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3213
    },
    {
      "epoch": 0.4370410660864835,
      "grad_norm": 0.4375,
      "learning_rate": 1.384163336302749e-05,
      "loss": 0.7078,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3214
    },
    {
      "epoch": 0.43717704650530326,
      "grad_norm": 0.44921875,
      "learning_rate": 1.3837249489077014e-05,
      "loss": 0.7925,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3215
    },
    {
      "epoch": 0.43731302692412294,
      "grad_norm": 0.65234375,
      "learning_rate": 1.3832864750163585e-05,
      "loss": 0.6415,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3216
    },
    {
      "epoch": 0.4374490073429426,
      "grad_norm": 0.47265625,
      "learning_rate": 1.3828479147275584e-05,
      "loss": 0.8081,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3217
    },
    {
      "epoch": 0.4375849877617623,
      "grad_norm": 0.322265625,
      "learning_rate": 1.3824092681401574e-05,
      "loss": 0.6753,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3218
    },
    {
      "epoch": 0.437720968180582,
      "grad_norm": 1.734375,
      "learning_rate": 1.3819705353530322e-05,
      "loss": 0.7401,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3219
    },
    {
      "epoch": 0.43785694859940166,
      "grad_norm": 0.373046875,
      "learning_rate": 1.3815317164650782e-05,
      "loss": 0.6624,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3220
    },
    {
      "epoch": 0.4379929290182214,
      "grad_norm": 0.296875,
      "learning_rate": 1.3810928115752104e-05,
      "loss": 0.5413,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3221
    },
    {
      "epoch": 0.4381289094370411,
      "grad_norm": 0.365234375,
      "learning_rate": 1.3806538207823641e-05,
      "loss": 0.6089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3222
    },
    {
      "epoch": 0.43826488985586076,
      "grad_norm": 0.455078125,
      "learning_rate": 1.3802147441854925e-05,
      "loss": 0.8076,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3223
    },
    {
      "epoch": 0.43840087027468044,
      "grad_norm": 0.31640625,
      "learning_rate": 1.379775581883569e-05,
      "loss": 0.6953,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3224
    },
    {
      "epoch": 0.4385368506935001,
      "grad_norm": 0.263671875,
      "learning_rate": 1.3793363339755863e-05,
      "loss": 0.4259,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3225
    },
    {
      "epoch": 0.4386728311123198,
      "grad_norm": 0.34375,
      "learning_rate": 1.378897000560556e-05,
      "loss": 0.658,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3226
    },
    {
      "epoch": 0.43880881153113954,
      "grad_norm": 0.4921875,
      "learning_rate": 1.3784575817375092e-05,
      "loss": 0.8607,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3227
    },
    {
      "epoch": 0.4389447919499592,
      "grad_norm": 0.32421875,
      "learning_rate": 1.3780180776054969e-05,
      "loss": 0.6738,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3228
    },
    {
      "epoch": 0.4390807723687789,
      "grad_norm": 0.58984375,
      "learning_rate": 1.3775784882635877e-05,
      "loss": 0.6865,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3229
    },
    {
      "epoch": 0.4392167527875986,
      "grad_norm": 0.396484375,
      "learning_rate": 1.3771388138108712e-05,
      "loss": 0.7373,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3230
    },
    {
      "epoch": 0.43935273320641827,
      "grad_norm": 0.57421875,
      "learning_rate": 1.3766990543464552e-05,
      "loss": 0.6479,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3231
    },
    {
      "epoch": 0.43948871362523795,
      "grad_norm": 0.37890625,
      "learning_rate": 1.3762592099694666e-05,
      "loss": 0.7673,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3232
    },
    {
      "epoch": 0.43962469404405763,
      "grad_norm": 0.390625,
      "learning_rate": 1.375819280779052e-05,
      "loss": 0.7567,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3233
    },
    {
      "epoch": 0.43976067446287737,
      "grad_norm": 0.56640625,
      "learning_rate": 1.3753792668743766e-05,
      "loss": 0.8455,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3234
    },
    {
      "epoch": 0.43989665488169705,
      "grad_norm": 0.3828125,
      "learning_rate": 1.3749391683546252e-05,
      "loss": 0.6899,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3235
    },
    {
      "epoch": 0.44003263530051673,
      "grad_norm": 0.462890625,
      "learning_rate": 1.3744989853190012e-05,
      "loss": 0.832,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3236
    },
    {
      "epoch": 0.4401686157193364,
      "grad_norm": 0.32421875,
      "learning_rate": 1.3740587178667273e-05,
      "loss": 0.6211,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3237
    },
    {
      "epoch": 0.4403045961381561,
      "grad_norm": 0.3828125,
      "learning_rate": 1.3736183660970449e-05,
      "loss": 0.5557,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3238
    },
    {
      "epoch": 0.44044057655697577,
      "grad_norm": 0.482421875,
      "learning_rate": 1.373177930109215e-05,
      "loss": 0.6654,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3239
    },
    {
      "epoch": 0.4405765569757955,
      "grad_norm": 0.462890625,
      "learning_rate": 1.3727374100025173e-05,
      "loss": 0.7804,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3240
    },
    {
      "epoch": 0.4407125373946152,
      "grad_norm": 0.56640625,
      "learning_rate": 1.37229680587625e-05,
      "loss": 0.7345,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3241
    },
    {
      "epoch": 0.44084851781343487,
      "grad_norm": 0.56640625,
      "learning_rate": 1.3718561178297311e-05,
      "loss": 0.6986,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3242
    },
    {
      "epoch": 0.44098449823225455,
      "grad_norm": 0.453125,
      "learning_rate": 1.371415345962297e-05,
      "loss": 0.8136,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3243
    },
    {
      "epoch": 0.44112047865107423,
      "grad_norm": 0.451171875,
      "learning_rate": 1.3709744903733027e-05,
      "loss": 0.8169,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3244
    },
    {
      "epoch": 0.4412564590698939,
      "grad_norm": 0.294921875,
      "learning_rate": 1.3705335511621229e-05,
      "loss": 0.5054,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3245
    },
    {
      "epoch": 0.44139243948871365,
      "grad_norm": 0.34375,
      "learning_rate": 1.3700925284281504e-05,
      "loss": 0.6803,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3246
    },
    {
      "epoch": 0.44152841990753333,
      "grad_norm": 0.28515625,
      "learning_rate": 1.3696514222707972e-05,
      "loss": 0.6169,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3247
    },
    {
      "epoch": 0.441664400326353,
      "grad_norm": 0.494140625,
      "learning_rate": 1.3692102327894942e-05,
      "loss": 0.7635,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3248
    },
    {
      "epoch": 0.4418003807451727,
      "grad_norm": 0.58203125,
      "learning_rate": 1.3687689600836906e-05,
      "loss": 0.9323,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3249
    },
    {
      "epoch": 0.4419363611639924,
      "grad_norm": 0.337890625,
      "learning_rate": 1.3683276042528548e-05,
      "loss": 0.6141,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3250
    },
    {
      "epoch": 0.44207234158281206,
      "grad_norm": 0.69921875,
      "learning_rate": 1.3678861653964738e-05,
      "loss": 0.6899,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3251
    },
    {
      "epoch": 0.44220832200163174,
      "grad_norm": 0.37890625,
      "learning_rate": 1.3674446436140534e-05,
      "loss": 0.6226,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3252
    },
    {
      "epoch": 0.4423443024204515,
      "grad_norm": 0.345703125,
      "learning_rate": 1.367003039005118e-05,
      "loss": 0.7049,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3253
    },
    {
      "epoch": 0.44248028283927116,
      "grad_norm": 0.2265625,
      "learning_rate": 1.3665613516692103e-05,
      "loss": 0.4149,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3254
    },
    {
      "epoch": 0.44261626325809084,
      "grad_norm": 0.80078125,
      "learning_rate": 1.3661195817058923e-05,
      "loss": 0.7856,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3255
    },
    {
      "epoch": 0.4427522436769105,
      "grad_norm": 0.2373046875,
      "learning_rate": 1.3656777292147445e-05,
      "loss": 0.445,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3256
    },
    {
      "epoch": 0.4428882240957302,
      "grad_norm": 0.609375,
      "learning_rate": 1.3652357942953658e-05,
      "loss": 0.7466,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3257
    },
    {
      "epoch": 0.4430242045145499,
      "grad_norm": 0.412109375,
      "learning_rate": 1.3647937770473739e-05,
      "loss": 0.66,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3258
    },
    {
      "epoch": 0.4431601849333696,
      "grad_norm": 0.447265625,
      "learning_rate": 1.3643516775704041e-05,
      "loss": 0.6603,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3259
    },
    {
      "epoch": 0.4432961653521893,
      "grad_norm": 0.361328125,
      "learning_rate": 1.3639094959641122e-05,
      "loss": 0.5946,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3260
    },
    {
      "epoch": 0.443432145771009,
      "grad_norm": 0.3515625,
      "learning_rate": 1.3634672323281702e-05,
      "loss": 0.5243,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3261
    },
    {
      "epoch": 0.44356812618982866,
      "grad_norm": 0.52734375,
      "learning_rate": 1.3630248867622705e-05,
      "loss": 0.5692,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3262
    },
    {
      "epoch": 0.44370410660864834,
      "grad_norm": 0.431640625,
      "learning_rate": 1.362582459366123e-05,
      "loss": 0.5273,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3263
    },
    {
      "epoch": 0.443840087027468,
      "grad_norm": 0.353515625,
      "learning_rate": 1.362139950239456e-05,
      "loss": 0.4847,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3264
    },
    {
      "epoch": 0.44397606744628776,
      "grad_norm": 0.73046875,
      "learning_rate": 1.361697359482017e-05,
      "loss": 0.5187,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3265
    },
    {
      "epoch": 0.44411204786510744,
      "grad_norm": 0.86328125,
      "learning_rate": 1.3612546871935705e-05,
      "loss": 0.7987,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3266
    },
    {
      "epoch": 0.4442480282839271,
      "grad_norm": 0.388671875,
      "learning_rate": 1.3608119334739011e-05,
      "loss": 0.6633,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3267
    },
    {
      "epoch": 0.4443840087027468,
      "grad_norm": 0.2294921875,
      "learning_rate": 1.3603690984228106e-05,
      "loss": 0.3774,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3268
    },
    {
      "epoch": 0.4445199891215665,
      "grad_norm": 0.4453125,
      "learning_rate": 1.3599261821401195e-05,
      "loss": 0.6763,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3269
    },
    {
      "epoch": 0.44465596954038616,
      "grad_norm": 0.39453125,
      "learning_rate": 1.3594831847256664e-05,
      "loss": 0.8634,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3270
    },
    {
      "epoch": 0.4447919499592059,
      "grad_norm": 0.283203125,
      "learning_rate": 1.3590401062793084e-05,
      "loss": 0.5978,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3271
    },
    {
      "epoch": 0.4449279303780256,
      "grad_norm": 0.470703125,
      "learning_rate": 1.3585969469009206e-05,
      "loss": 0.6686,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3272
    },
    {
      "epoch": 0.44506391079684526,
      "grad_norm": 0.34375,
      "learning_rate": 1.3581537066903966e-05,
      "loss": 0.6064,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3273
    },
    {
      "epoch": 0.44519989121566494,
      "grad_norm": 0.5078125,
      "learning_rate": 1.3577103857476488e-05,
      "loss": 0.8535,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3274
    },
    {
      "epoch": 0.4453358716344846,
      "grad_norm": 0.37890625,
      "learning_rate": 1.3572669841726062e-05,
      "loss": 0.71,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3275
    },
    {
      "epoch": 0.4454718520533043,
      "grad_norm": 0.578125,
      "learning_rate": 1.3568235020652178e-05,
      "loss": 0.7422,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3276
    },
    {
      "epoch": 0.445607832472124,
      "grad_norm": 0.5078125,
      "learning_rate": 1.3563799395254494e-05,
      "loss": 0.7619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3277
    },
    {
      "epoch": 0.4457438128909437,
      "grad_norm": 0.3671875,
      "learning_rate": 1.355936296653285e-05,
      "loss": 0.6982,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3278
    },
    {
      "epoch": 0.4458797933097634,
      "grad_norm": 0.373046875,
      "learning_rate": 1.3554925735487281e-05,
      "loss": 0.6361,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3279
    },
    {
      "epoch": 0.4460157737285831,
      "grad_norm": 0.65625,
      "learning_rate": 1.3550487703117987e-05,
      "loss": 0.6579,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3280
    },
    {
      "epoch": 0.44615175414740277,
      "grad_norm": 0.30078125,
      "learning_rate": 1.3546048870425356e-05,
      "loss": 0.5132,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3281
    },
    {
      "epoch": 0.44628773456622245,
      "grad_norm": 0.35546875,
      "learning_rate": 1.3541609238409956e-05,
      "loss": 0.5277,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3282
    },
    {
      "epoch": 0.44642371498504213,
      "grad_norm": 0.74609375,
      "learning_rate": 1.3537168808072534e-05,
      "loss": 0.7974,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3283
    },
    {
      "epoch": 0.44655969540386187,
      "grad_norm": 0.55859375,
      "learning_rate": 1.3532727580414018e-05,
      "loss": 0.6515,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3284
    },
    {
      "epoch": 0.44669567582268155,
      "grad_norm": 0.2890625,
      "learning_rate": 1.3528285556435512e-05,
      "loss": 0.57,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3285
    },
    {
      "epoch": 0.44683165624150123,
      "grad_norm": 0.369140625,
      "learning_rate": 1.3523842737138306e-05,
      "loss": 0.6042,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3286
    },
    {
      "epoch": 0.4469676366603209,
      "grad_norm": 0.76171875,
      "learning_rate": 1.3519399123523862e-05,
      "loss": 0.5876,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3287
    },
    {
      "epoch": 0.4471036170791406,
      "grad_norm": 0.443359375,
      "learning_rate": 1.3514954716593828e-05,
      "loss": 0.7323,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3288
    },
    {
      "epoch": 0.4472395974979603,
      "grad_norm": 0.435546875,
      "learning_rate": 1.3510509517350028e-05,
      "loss": 0.576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3289
    },
    {
      "epoch": 0.44737557791678,
      "grad_norm": 0.515625,
      "learning_rate": 1.3506063526794462e-05,
      "loss": 0.738,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3290
    },
    {
      "epoch": 0.4475115583355997,
      "grad_norm": 0.337890625,
      "learning_rate": 1.3501616745929314e-05,
      "loss": 0.5824,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3291
    },
    {
      "epoch": 0.44764753875441937,
      "grad_norm": 0.60546875,
      "learning_rate": 1.3497169175756936e-05,
      "loss": 0.6276,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3292
    },
    {
      "epoch": 0.44778351917323905,
      "grad_norm": 0.44140625,
      "learning_rate": 1.3492720817279872e-05,
      "loss": 0.6252,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3293
    },
    {
      "epoch": 0.44791949959205873,
      "grad_norm": 0.462890625,
      "learning_rate": 1.3488271671500832e-05,
      "loss": 0.7464,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3294
    },
    {
      "epoch": 0.4480554800108784,
      "grad_norm": 0.45703125,
      "learning_rate": 1.348382173942271e-05,
      "loss": 0.8244,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3295
    },
    {
      "epoch": 0.4481914604296981,
      "grad_norm": 0.2578125,
      "learning_rate": 1.3479371022048577e-05,
      "loss": 0.382,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3296
    },
    {
      "epoch": 0.44832744084851783,
      "grad_norm": 0.28125,
      "learning_rate": 1.3474919520381673e-05,
      "loss": 0.5392,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3297
    },
    {
      "epoch": 0.4484634212673375,
      "grad_norm": 0.474609375,
      "learning_rate": 1.3470467235425426e-05,
      "loss": 0.9256,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3298
    },
    {
      "epoch": 0.4485994016861572,
      "grad_norm": 0.37890625,
      "learning_rate": 1.3466014168183434e-05,
      "loss": 0.6237,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3299
    },
    {
      "epoch": 0.4487353821049769,
      "grad_norm": 0.45703125,
      "learning_rate": 1.3461560319659472e-05,
      "loss": 0.6963,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3300
    },
    {
      "epoch": 0.44887136252379656,
      "grad_norm": 0.322265625,
      "learning_rate": 1.3457105690857494e-05,
      "loss": 0.5807,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3301
    },
    {
      "epoch": 0.44900734294261624,
      "grad_norm": 0.337890625,
      "learning_rate": 1.3452650282781625e-05,
      "loss": 0.606,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3302
    },
    {
      "epoch": 0.449143323361436,
      "grad_norm": 0.388671875,
      "learning_rate": 1.3448194096436171e-05,
      "loss": 0.6914,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3303
    },
    {
      "epoch": 0.44927930378025566,
      "grad_norm": 0.53515625,
      "learning_rate": 1.3443737132825609e-05,
      "loss": 0.5555,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3304
    },
    {
      "epoch": 0.44941528419907534,
      "grad_norm": 0.328125,
      "learning_rate": 1.3439279392954594e-05,
      "loss": 0.5967,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3305
    },
    {
      "epoch": 0.449551264617895,
      "grad_norm": 0.33984375,
      "learning_rate": 1.3434820877827957e-05,
      "loss": 0.5793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3306
    },
    {
      "epoch": 0.4496872450367147,
      "grad_norm": 0.36328125,
      "learning_rate": 1.3430361588450696e-05,
      "loss": 0.6374,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3307
    },
    {
      "epoch": 0.4498232254555344,
      "grad_norm": 0.28515625,
      "learning_rate": 1.3425901525827993e-05,
      "loss": 0.5117,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3308
    },
    {
      "epoch": 0.4499592058743541,
      "grad_norm": 0.271484375,
      "learning_rate": 1.3421440690965203e-05,
      "loss": 0.432,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3309
    },
    {
      "epoch": 0.4500951862931738,
      "grad_norm": 0.27734375,
      "learning_rate": 1.3416979084867851e-05,
      "loss": 0.562,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3310
    },
    {
      "epoch": 0.4502311667119935,
      "grad_norm": 0.546875,
      "learning_rate": 1.3412516708541637e-05,
      "loss": 0.6413,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3311
    },
    {
      "epoch": 0.45036714713081316,
      "grad_norm": 0.8828125,
      "learning_rate": 1.3408053562992433e-05,
      "loss": 0.779,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3312
    },
    {
      "epoch": 0.45050312754963284,
      "grad_norm": 0.298828125,
      "learning_rate": 1.3403589649226292e-05,
      "loss": 0.5223,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3313
    },
    {
      "epoch": 0.4506391079684525,
      "grad_norm": 1.078125,
      "learning_rate": 1.3399124968249428e-05,
      "loss": 0.5498,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3314
    },
    {
      "epoch": 0.45077508838727226,
      "grad_norm": 0.59375,
      "learning_rate": 1.3394659521068239e-05,
      "loss": 0.6901,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3315
    },
    {
      "epoch": 0.45091106880609194,
      "grad_norm": 0.62890625,
      "learning_rate": 1.3390193308689293e-05,
      "loss": 0.8263,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3316
    },
    {
      "epoch": 0.4510470492249116,
      "grad_norm": 0.359375,
      "learning_rate": 1.3385726332119324e-05,
      "loss": 0.6478,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3317
    },
    {
      "epoch": 0.4511830296437313,
      "grad_norm": 0.474609375,
      "learning_rate": 1.3381258592365245e-05,
      "loss": 0.6605,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3318
    },
    {
      "epoch": 0.451319010062551,
      "grad_norm": 0.3359375,
      "learning_rate": 1.3376790090434141e-05,
      "loss": 0.6076,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3319
    },
    {
      "epoch": 0.45145499048137067,
      "grad_norm": 0.63671875,
      "learning_rate": 1.3372320827333267e-05,
      "loss": 0.6628,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3320
    },
    {
      "epoch": 0.45159097090019035,
      "grad_norm": 0.3203125,
      "learning_rate": 1.3367850804070046e-05,
      "loss": 0.7681,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3321
    },
    {
      "epoch": 0.4517269513190101,
      "grad_norm": 0.59375,
      "learning_rate": 1.336338002165208e-05,
      "loss": 0.3981,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3322
    },
    {
      "epoch": 0.45186293173782976,
      "grad_norm": 0.396484375,
      "learning_rate": 1.3358908481087133e-05,
      "loss": 0.7611,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3323
    },
    {
      "epoch": 0.45199891215664945,
      "grad_norm": 0.46875,
      "learning_rate": 1.3354436183383152e-05,
      "loss": 0.6867,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3324
    },
    {
      "epoch": 0.4521348925754691,
      "grad_norm": 0.87109375,
      "learning_rate": 1.3349963129548243e-05,
      "loss": 0.6686,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3325
    },
    {
      "epoch": 0.4522708729942888,
      "grad_norm": 0.388671875,
      "learning_rate": 1.3345489320590686e-05,
      "loss": 0.7262,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3326
    },
    {
      "epoch": 0.4524068534131085,
      "grad_norm": 0.52734375,
      "learning_rate": 1.3341014757518935e-05,
      "loss": 0.5101,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3327
    },
    {
      "epoch": 0.4525428338319282,
      "grad_norm": 0.3359375,
      "learning_rate": 1.333653944134161e-05,
      "loss": 0.5275,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3328
    },
    {
      "epoch": 0.4526788142507479,
      "grad_norm": 0.359375,
      "learning_rate": 1.33320633730675e-05,
      "loss": 0.494,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3329
    },
    {
      "epoch": 0.4528147946695676,
      "grad_norm": 0.388671875,
      "learning_rate": 1.332758655370557e-05,
      "loss": 0.6659,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3330
    },
    {
      "epoch": 0.45295077508838727,
      "grad_norm": 0.28515625,
      "learning_rate": 1.3323108984264948e-05,
      "loss": 0.5129,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3331
    },
    {
      "epoch": 0.45308675550720695,
      "grad_norm": 0.26171875,
      "learning_rate": 1.331863066575493e-05,
      "loss": 0.4533,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3332
    },
    {
      "epoch": 0.45322273592602663,
      "grad_norm": 0.466796875,
      "learning_rate": 1.3314151599184988e-05,
      "loss": 0.4948,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3333
    },
    {
      "epoch": 0.45335871634484637,
      "grad_norm": 0.265625,
      "learning_rate": 1.3309671785564757e-05,
      "loss": 0.4624,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3334
    },
    {
      "epoch": 0.45349469676366605,
      "grad_norm": 0.388671875,
      "learning_rate": 1.3305191225904039e-05,
      "loss": 0.6467,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3335
    },
    {
      "epoch": 0.45363067718248573,
      "grad_norm": 0.51953125,
      "learning_rate": 1.330070992121281e-05,
      "loss": 0.8799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3336
    },
    {
      "epoch": 0.4537666576013054,
      "grad_norm": 0.412109375,
      "learning_rate": 1.329622787250121e-05,
      "loss": 0.5452,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3337
    },
    {
      "epoch": 0.4539026380201251,
      "grad_norm": 0.58203125,
      "learning_rate": 1.3291745080779545e-05,
      "loss": 0.5687,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3338
    },
    {
      "epoch": 0.4540386184389448,
      "grad_norm": 0.4296875,
      "learning_rate": 1.3287261547058298e-05,
      "loss": 0.6997,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3339
    },
    {
      "epoch": 0.45417459885776446,
      "grad_norm": 0.361328125,
      "learning_rate": 1.3282777272348106e-05,
      "loss": 0.756,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3340
    },
    {
      "epoch": 0.4543105792765842,
      "grad_norm": 0.5546875,
      "learning_rate": 1.3278292257659783e-05,
      "loss": 0.7824,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3341
    },
    {
      "epoch": 0.4544465596954039,
      "grad_norm": 0.37890625,
      "learning_rate": 1.3273806504004304e-05,
      "loss": 0.6743,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3342
    },
    {
      "epoch": 0.45458254011422355,
      "grad_norm": 0.3984375,
      "learning_rate": 1.326932001239281e-05,
      "loss": 0.5205,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3343
    },
    {
      "epoch": 0.45471852053304324,
      "grad_norm": 0.462890625,
      "learning_rate": 1.3264832783836618e-05,
      "loss": 0.8091,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3344
    },
    {
      "epoch": 0.4548545009518629,
      "grad_norm": 0.4453125,
      "learning_rate": 1.3260344819347199e-05,
      "loss": 0.832,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3345
    },
    {
      "epoch": 0.4549904813706826,
      "grad_norm": 0.443359375,
      "learning_rate": 1.3255856119936197e-05,
      "loss": 0.8051,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3346
    },
    {
      "epoch": 0.45512646178950233,
      "grad_norm": 0.326171875,
      "learning_rate": 1.3251366686615419e-05,
      "loss": 0.6535,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3347
    },
    {
      "epoch": 0.455262442208322,
      "grad_norm": 0.353515625,
      "learning_rate": 1.3246876520396837e-05,
      "loss": 0.6903,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3348
    },
    {
      "epoch": 0.4553984226271417,
      "grad_norm": 0.353515625,
      "learning_rate": 1.3242385622292593e-05,
      "loss": 0.6276,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3349
    },
    {
      "epoch": 0.4555344030459614,
      "grad_norm": 0.271484375,
      "learning_rate": 1.3237893993314988e-05,
      "loss": 0.5812,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3350
    },
    {
      "epoch": 0.45567038346478106,
      "grad_norm": 0.2734375,
      "learning_rate": 1.3233401634476487e-05,
      "loss": 0.5187,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3351
    },
    {
      "epoch": 0.45580636388360074,
      "grad_norm": 0.419921875,
      "learning_rate": 1.3228908546789729e-05,
      "loss": 0.6667,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3352
    },
    {
      "epoch": 0.4559423443024205,
      "grad_norm": 0.34375,
      "learning_rate": 1.3224414731267508e-05,
      "loss": 0.5986,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3353
    },
    {
      "epoch": 0.45607832472124016,
      "grad_norm": 0.50390625,
      "learning_rate": 1.321992018892278e-05,
      "loss": 0.5542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3354
    },
    {
      "epoch": 0.45621430514005984,
      "grad_norm": 0.345703125,
      "learning_rate": 1.3215424920768678e-05,
      "loss": 0.5841,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3355
    },
    {
      "epoch": 0.4563502855588795,
      "grad_norm": 0.365234375,
      "learning_rate": 1.3210928927818487e-05,
      "loss": 0.6087,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3356
    },
    {
      "epoch": 0.4564862659776992,
      "grad_norm": 0.28125,
      "learning_rate": 1.3206432211085657e-05,
      "loss": 0.6559,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3357
    },
    {
      "epoch": 0.4566222463965189,
      "grad_norm": 0.330078125,
      "learning_rate": 1.3201934771583807e-05,
      "loss": 0.6193,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3358
    },
    {
      "epoch": 0.45675822681533856,
      "grad_norm": 0.392578125,
      "learning_rate": 1.319743661032671e-05,
      "loss": 0.5288,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3359
    },
    {
      "epoch": 0.4568942072341583,
      "grad_norm": 0.3203125,
      "learning_rate": 1.319293772832831e-05,
      "loss": 0.6074,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3360
    },
    {
      "epoch": 0.457030187652978,
      "grad_norm": 0.390625,
      "learning_rate": 1.3188438126602709e-05,
      "loss": 0.6575,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3361
    },
    {
      "epoch": 0.45716616807179766,
      "grad_norm": 0.2294921875,
      "learning_rate": 1.3183937806164174e-05,
      "loss": 0.3724,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3362
    },
    {
      "epoch": 0.45730214849061734,
      "grad_norm": 1.453125,
      "learning_rate": 1.3179436768027128e-05,
      "loss": 0.8695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3363
    },
    {
      "epoch": 0.457438128909437,
      "grad_norm": 0.39453125,
      "learning_rate": 1.317493501320616e-05,
      "loss": 0.6305,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3364
    },
    {
      "epoch": 0.4575741093282567,
      "grad_norm": 0.380859375,
      "learning_rate": 1.3170432542716028e-05,
      "loss": 0.6968,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3365
    },
    {
      "epoch": 0.45771008974707644,
      "grad_norm": 0.3515625,
      "learning_rate": 1.3165929357571636e-05,
      "loss": 0.6362,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3366
    },
    {
      "epoch": 0.4578460701658961,
      "grad_norm": 0.3828125,
      "learning_rate": 1.316142545878806e-05,
      "loss": 0.6834,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3367
    },
    {
      "epoch": 0.4579820505847158,
      "grad_norm": 0.375,
      "learning_rate": 1.3156920847380535e-05,
      "loss": 0.541,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3368
    },
    {
      "epoch": 0.4581180310035355,
      "grad_norm": 0.58203125,
      "learning_rate": 1.3152415524364454e-05,
      "loss": 0.7355,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3369
    },
    {
      "epoch": 0.45825401142235517,
      "grad_norm": 0.59375,
      "learning_rate": 1.314790949075537e-05,
      "loss": 0.7855,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3370
    },
    {
      "epoch": 0.45838999184117485,
      "grad_norm": 0.470703125,
      "learning_rate": 1.3143402747569003e-05,
      "loss": 0.849,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3371
    },
    {
      "epoch": 0.4585259722599946,
      "grad_norm": 0.40625,
      "learning_rate": 1.3138895295821224e-05,
      "loss": 0.7869,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3372
    },
    {
      "epoch": 0.45866195267881427,
      "grad_norm": 1.8828125,
      "learning_rate": 1.3134387136528069e-05,
      "loss": 0.91,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3373
    },
    {
      "epoch": 0.45879793309763395,
      "grad_norm": 0.310546875,
      "learning_rate": 1.3129878270705732e-05,
      "loss": 0.5207,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3374
    },
    {
      "epoch": 0.45893391351645363,
      "grad_norm": 0.345703125,
      "learning_rate": 1.3125368699370567e-05,
      "loss": 0.569,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3375
    },
    {
      "epoch": 0.4590698939352733,
      "grad_norm": 0.349609375,
      "learning_rate": 1.3120858423539086e-05,
      "loss": 0.7427,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3376
    },
    {
      "epoch": 0.459205874354093,
      "grad_norm": 0.31640625,
      "learning_rate": 1.3116347444227961e-05,
      "loss": 0.5851,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3377
    },
    {
      "epoch": 0.4593418547729127,
      "grad_norm": 0.263671875,
      "learning_rate": 1.311183576245402e-05,
      "loss": 0.6541,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3378
    },
    {
      "epoch": 0.4594778351917324,
      "grad_norm": 0.35546875,
      "learning_rate": 1.3107323379234254e-05,
      "loss": 0.5882,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3379
    },
    {
      "epoch": 0.4596138156105521,
      "grad_norm": 0.43359375,
      "learning_rate": 1.3102810295585808e-05,
      "loss": 0.5355,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3380
    },
    {
      "epoch": 0.45974979602937177,
      "grad_norm": 0.498046875,
      "learning_rate": 1.3098296512525986e-05,
      "loss": 0.7643,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3381
    },
    {
      "epoch": 0.45988577644819145,
      "grad_norm": 0.42578125,
      "learning_rate": 1.3093782031072249e-05,
      "loss": 0.5396,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3382
    },
    {
      "epoch": 0.46002175686701113,
      "grad_norm": 0.349609375,
      "learning_rate": 1.3089266852242217e-05,
      "loss": 0.7212,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3383
    },
    {
      "epoch": 0.4601577372858308,
      "grad_norm": 0.55078125,
      "learning_rate": 1.3084750977053669e-05,
      "loss": 0.5926,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3384
    },
    {
      "epoch": 0.46029371770465055,
      "grad_norm": 0.455078125,
      "learning_rate": 1.3080234406524534e-05,
      "loss": 0.7699,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3385
    },
    {
      "epoch": 0.46042969812347023,
      "grad_norm": 0.43359375,
      "learning_rate": 1.3075717141672905e-05,
      "loss": 0.7285,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3386
    },
    {
      "epoch": 0.4605656785422899,
      "grad_norm": 0.384765625,
      "learning_rate": 1.3071199183517028e-05,
      "loss": 0.626,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3387
    },
    {
      "epoch": 0.4607016589611096,
      "grad_norm": 0.318359375,
      "learning_rate": 1.3066680533075312e-05,
      "loss": 0.6533,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3388
    },
    {
      "epoch": 0.4608376393799293,
      "grad_norm": 0.373046875,
      "learning_rate": 1.3062161191366303e-05,
      "loss": 0.6657,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3389
    },
    {
      "epoch": 0.46097361979874896,
      "grad_norm": 0.376953125,
      "learning_rate": 1.3057641159408726e-05,
      "loss": 0.6825,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3390
    },
    {
      "epoch": 0.4611096002175687,
      "grad_norm": 0.490234375,
      "learning_rate": 1.3053120438221449e-05,
      "loss": 0.7477,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3391
    },
    {
      "epoch": 0.4612455806363884,
      "grad_norm": 0.5703125,
      "learning_rate": 1.30485990288235e-05,
      "loss": 0.7694,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3392
    },
    {
      "epoch": 0.46138156105520806,
      "grad_norm": 0.44140625,
      "learning_rate": 1.3044076932234054e-05,
      "loss": 0.59,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3393
    },
    {
      "epoch": 0.46151754147402774,
      "grad_norm": 0.330078125,
      "learning_rate": 1.3039554149472449e-05,
      "loss": 0.6675,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3394
    },
    {
      "epoch": 0.4616535218928474,
      "grad_norm": 0.26171875,
      "learning_rate": 1.3035030681558178e-05,
      "loss": 0.4139,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3395
    },
    {
      "epoch": 0.4617895023116671,
      "grad_norm": 0.408203125,
      "learning_rate": 1.3030506529510887e-05,
      "loss": 0.7949,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3396
    },
    {
      "epoch": 0.46192548273048684,
      "grad_norm": 0.68359375,
      "learning_rate": 1.3025981694350371e-05,
      "loss": 0.7352,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3397
    },
    {
      "epoch": 0.4620614631493065,
      "grad_norm": 0.3828125,
      "learning_rate": 1.3021456177096585e-05,
      "loss": 0.7213,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3398
    },
    {
      "epoch": 0.4621974435681262,
      "grad_norm": 0.26953125,
      "learning_rate": 1.3016929978769637e-05,
      "loss": 0.542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3399
    },
    {
      "epoch": 0.4623334239869459,
      "grad_norm": 0.671875,
      "learning_rate": 1.3012403100389785e-05,
      "loss": 0.6346,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3400
    },
    {
      "epoch": 0.46246940440576556,
      "grad_norm": 0.578125,
      "learning_rate": 1.3007875542977448e-05,
      "loss": 0.5745,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3401
    },
    {
      "epoch": 0.46260538482458524,
      "grad_norm": 0.5859375,
      "learning_rate": 1.3003347307553187e-05,
      "loss": 0.5815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3402
    },
    {
      "epoch": 0.4627413652434049,
      "grad_norm": 0.32421875,
      "learning_rate": 1.2998818395137725e-05,
      "loss": 0.687,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3403
    },
    {
      "epoch": 0.46287734566222466,
      "grad_norm": 0.33203125,
      "learning_rate": 1.2994288806751932e-05,
      "loss": 0.6507,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3404
    },
    {
      "epoch": 0.46301332608104434,
      "grad_norm": 0.3046875,
      "learning_rate": 1.2989758543416832e-05,
      "loss": 0.576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3405
    },
    {
      "epoch": 0.463149306499864,
      "grad_norm": 0.5078125,
      "learning_rate": 1.2985227606153604e-05,
      "loss": 0.5977,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3406
    },
    {
      "epoch": 0.4632852869186837,
      "grad_norm": 0.734375,
      "learning_rate": 1.2980695995983576e-05,
      "loss": 0.7155,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3407
    },
    {
      "epoch": 0.4634212673375034,
      "grad_norm": 0.5078125,
      "learning_rate": 1.297616371392823e-05,
      "loss": 0.9053,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3408
    },
    {
      "epoch": 0.46355724775632307,
      "grad_norm": 0.486328125,
      "learning_rate": 1.2971630761009193e-05,
      "loss": 0.9463,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3409
    },
    {
      "epoch": 0.4636932281751428,
      "grad_norm": 0.46484375,
      "learning_rate": 1.2967097138248256e-05,
      "loss": 0.7352,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3410
    },
    {
      "epoch": 0.4638292085939625,
      "grad_norm": 0.3125,
      "learning_rate": 1.2962562846667344e-05,
      "loss": 0.6576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3411
    },
    {
      "epoch": 0.46396518901278216,
      "grad_norm": 0.341796875,
      "learning_rate": 1.2958027887288546e-05,
      "loss": 0.5643,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3412
    },
    {
      "epoch": 0.46410116943160185,
      "grad_norm": 0.369140625,
      "learning_rate": 1.2953492261134097e-05,
      "loss": 0.7072,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3413
    },
    {
      "epoch": 0.4642371498504215,
      "grad_norm": 0.390625,
      "learning_rate": 1.2948955969226384e-05,
      "loss": 0.6997,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3414
    },
    {
      "epoch": 0.4643731302692412,
      "grad_norm": 0.482421875,
      "learning_rate": 1.2944419012587941e-05,
      "loss": 0.7284,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3415
    },
    {
      "epoch": 0.46450911068806094,
      "grad_norm": 0.423828125,
      "learning_rate": 1.2939881392241453e-05,
      "loss": 0.7165,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3416
    },
    {
      "epoch": 0.4646450911068806,
      "grad_norm": 0.44140625,
      "learning_rate": 1.293534310920976e-05,
      "loss": 0.5685,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3417
    },
    {
      "epoch": 0.4647810715257003,
      "grad_norm": 0.322265625,
      "learning_rate": 1.2930804164515837e-05,
      "loss": 0.3879,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3418
    },
    {
      "epoch": 0.46491705194452,
      "grad_norm": 0.61328125,
      "learning_rate": 1.2926264559182825e-05,
      "loss": 0.7778,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3419
    },
    {
      "epoch": 0.46505303236333967,
      "grad_norm": 1.1484375,
      "learning_rate": 1.2921724294234005e-05,
      "loss": 0.8136,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3420
    },
    {
      "epoch": 0.46518901278215935,
      "grad_norm": 0.43359375,
      "learning_rate": 1.291718337069281e-05,
      "loss": 0.6374,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3421
    },
    {
      "epoch": 0.46532499320097903,
      "grad_norm": 0.384765625,
      "learning_rate": 1.2912641789582813e-05,
      "loss": 0.724,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3422
    },
    {
      "epoch": 0.46546097361979877,
      "grad_norm": 0.41015625,
      "learning_rate": 1.290809955192775e-05,
      "loss": 0.6989,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3423
    },
    {
      "epoch": 0.46559695403861845,
      "grad_norm": 0.57421875,
      "learning_rate": 1.2903556658751493e-05,
      "loss": 0.5936,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3424
    },
    {
      "epoch": 0.46573293445743813,
      "grad_norm": 0.6640625,
      "learning_rate": 1.2899013111078065e-05,
      "loss": 0.8094,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3425
    },
    {
      "epoch": 0.4658689148762578,
      "grad_norm": 0.44921875,
      "learning_rate": 1.2894468909931636e-05,
      "loss": 0.765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3426
    },
    {
      "epoch": 0.4660048952950775,
      "grad_norm": 1.0234375,
      "learning_rate": 1.2889924056336531e-05,
      "loss": 0.6065,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3427
    },
    {
      "epoch": 0.4661408757138972,
      "grad_norm": 0.578125,
      "learning_rate": 1.288537855131721e-05,
      "loss": 0.5422,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3428
    },
    {
      "epoch": 0.4662768561327169,
      "grad_norm": 0.6796875,
      "learning_rate": 1.2880832395898285e-05,
      "loss": 0.733,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3429
    },
    {
      "epoch": 0.4664128365515366,
      "grad_norm": 0.365234375,
      "learning_rate": 1.2876285591104519e-05,
      "loss": 0.5693,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3430
    },
    {
      "epoch": 0.4665488169703563,
      "grad_norm": 0.7890625,
      "learning_rate": 1.2871738137960811e-05,
      "loss": 0.6823,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3431
    },
    {
      "epoch": 0.46668479738917595,
      "grad_norm": 0.32421875,
      "learning_rate": 1.2867190037492221e-05,
      "loss": 0.6053,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3432
    },
    {
      "epoch": 0.46682077780799563,
      "grad_norm": 0.306640625,
      "learning_rate": 1.286264129072394e-05,
      "loss": 0.5692,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3433
    },
    {
      "epoch": 0.4669567582268153,
      "grad_norm": 0.640625,
      "learning_rate": 1.2858091898681313e-05,
      "loss": 0.6858,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3434
    },
    {
      "epoch": 0.46709273864563505,
      "grad_norm": 0.29296875,
      "learning_rate": 1.285354186238983e-05,
      "loss": 0.4582,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3435
    },
    {
      "epoch": 0.46722871906445473,
      "grad_norm": 0.55859375,
      "learning_rate": 1.2848991182875124e-05,
      "loss": 0.5817,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3436
    },
    {
      "epoch": 0.4673646994832744,
      "grad_norm": 0.32421875,
      "learning_rate": 1.2844439861162972e-05,
      "loss": 0.6484,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3437
    },
    {
      "epoch": 0.4675006799020941,
      "grad_norm": 0.7734375,
      "learning_rate": 1.2839887898279298e-05,
      "loss": 0.7653,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3438
    },
    {
      "epoch": 0.4676366603209138,
      "grad_norm": 0.390625,
      "learning_rate": 1.2835335295250172e-05,
      "loss": 0.674,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3439
    },
    {
      "epoch": 0.46777264073973346,
      "grad_norm": 0.3359375,
      "learning_rate": 1.2830782053101807e-05,
      "loss": 0.6182,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3440
    },
    {
      "epoch": 0.4679086211585532,
      "grad_norm": 0.30078125,
      "learning_rate": 1.2826228172860555e-05,
      "loss": 0.6004,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3441
    },
    {
      "epoch": 0.4680446015773729,
      "grad_norm": 0.427734375,
      "learning_rate": 1.2821673655552922e-05,
      "loss": 0.652,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3442
    },
    {
      "epoch": 0.46818058199619256,
      "grad_norm": 0.49609375,
      "learning_rate": 1.2817118502205547e-05,
      "loss": 0.8846,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3443
    },
    {
      "epoch": 0.46831656241501224,
      "grad_norm": 0.5,
      "learning_rate": 1.281256271384522e-05,
      "loss": 0.6413,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3444
    },
    {
      "epoch": 0.4684525428338319,
      "grad_norm": 0.45703125,
      "learning_rate": 1.2808006291498872e-05,
      "loss": 0.7471,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3445
    },
    {
      "epoch": 0.4685885232526516,
      "grad_norm": 0.48046875,
      "learning_rate": 1.2803449236193573e-05,
      "loss": 0.569,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3446
    },
    {
      "epoch": 0.4687245036714713,
      "grad_norm": 0.470703125,
      "learning_rate": 1.2798891548956543e-05,
      "loss": 0.3838,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3447
    },
    {
      "epoch": 0.468860484090291,
      "grad_norm": 0.330078125,
      "learning_rate": 1.2794333230815137e-05,
      "loss": 0.5667,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3448
    },
    {
      "epoch": 0.4689964645091107,
      "grad_norm": 0.326171875,
      "learning_rate": 1.2789774282796857e-05,
      "loss": 0.6711,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3449
    },
    {
      "epoch": 0.4691324449279304,
      "grad_norm": 0.19921875,
      "learning_rate": 1.2785214705929345e-05,
      "loss": 0.3374,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3450
    },
    {
      "epoch": 0.46926842534675006,
      "grad_norm": 0.365234375,
      "learning_rate": 1.2780654501240387e-05,
      "loss": 0.7368,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3451
    },
    {
      "epoch": 0.46940440576556974,
      "grad_norm": 0.6484375,
      "learning_rate": 1.2776093669757907e-05,
      "loss": 0.6839,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3452
    },
    {
      "epoch": 0.4695403861843894,
      "grad_norm": 0.34375,
      "learning_rate": 1.2771532212509974e-05,
      "loss": 0.6045,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3453
    },
    {
      "epoch": 0.46967636660320916,
      "grad_norm": 0.373046875,
      "learning_rate": 1.2766970130524797e-05,
      "loss": 0.5672,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3454
    },
    {
      "epoch": 0.46981234702202884,
      "grad_norm": 0.3359375,
      "learning_rate": 1.276240742483072e-05,
      "loss": 0.481,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3455
    },
    {
      "epoch": 0.4699483274408485,
      "grad_norm": 0.2734375,
      "learning_rate": 1.2757844096456238e-05,
      "loss": 0.4946,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3456
    },
    {
      "epoch": 0.4700843078596682,
      "grad_norm": 0.412109375,
      "learning_rate": 1.2753280146429978e-05,
      "loss": 0.693,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3457
    },
    {
      "epoch": 0.4702202882784879,
      "grad_norm": 0.52734375,
      "learning_rate": 1.274871557578071e-05,
      "loss": 0.6838,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3458
    },
    {
      "epoch": 0.47035626869730757,
      "grad_norm": 0.318359375,
      "learning_rate": 1.2744150385537346e-05,
      "loss": 0.6693,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3459
    },
    {
      "epoch": 0.4704922491161273,
      "grad_norm": 0.2470703125,
      "learning_rate": 1.2739584576728935e-05,
      "loss": 0.55,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3460
    },
    {
      "epoch": 0.470628229534947,
      "grad_norm": 0.375,
      "learning_rate": 1.2735018150384667e-05,
      "loss": 0.6353,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3461
    },
    {
      "epoch": 0.47076420995376667,
      "grad_norm": 0.337890625,
      "learning_rate": 1.273045110753387e-05,
      "loss": 0.6527,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3462
    },
    {
      "epoch": 0.47090019037258635,
      "grad_norm": 0.62890625,
      "learning_rate": 1.2725883449206009e-05,
      "loss": 0.8512,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3463
    },
    {
      "epoch": 0.47103617079140603,
      "grad_norm": 0.32421875,
      "learning_rate": 1.2721315176430692e-05,
      "loss": 0.708,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3464
    },
    {
      "epoch": 0.4711721512102257,
      "grad_norm": 0.267578125,
      "learning_rate": 1.2716746290237664e-05,
      "loss": 0.5679,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3465
    },
    {
      "epoch": 0.4713081316290454,
      "grad_norm": 0.34375,
      "learning_rate": 1.2712176791656807e-05,
      "loss": 0.6888,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3466
    },
    {
      "epoch": 0.4714441120478651,
      "grad_norm": 0.33203125,
      "learning_rate": 1.2707606681718143e-05,
      "loss": 0.6815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3467
    },
    {
      "epoch": 0.4715800924666848,
      "grad_norm": 0.41796875,
      "learning_rate": 1.2703035961451829e-05,
      "loss": 0.6921,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3468
    },
    {
      "epoch": 0.4717160728855045,
      "grad_norm": 0.396484375,
      "learning_rate": 1.2698464631888161e-05,
      "loss": 0.6681,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3469
    },
    {
      "epoch": 0.47185205330432417,
      "grad_norm": 0.40234375,
      "learning_rate": 1.2693892694057573e-05,
      "loss": 0.6919,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3470
    },
    {
      "epoch": 0.47198803372314385,
      "grad_norm": 0.314453125,
      "learning_rate": 1.2689320148990636e-05,
      "loss": 0.5942,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3471
    },
    {
      "epoch": 0.47212401414196353,
      "grad_norm": 1.3125,
      "learning_rate": 1.2684746997718059e-05,
      "loss": 0.7541,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3472
    },
    {
      "epoch": 0.47225999456078327,
      "grad_norm": 0.451171875,
      "learning_rate": 1.2680173241270683e-05,
      "loss": 0.5527,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3473
    },
    {
      "epoch": 0.47239597497960295,
      "grad_norm": 0.62890625,
      "learning_rate": 1.267559888067949e-05,
      "loss": 0.6144,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3474
    },
    {
      "epoch": 0.47253195539842263,
      "grad_norm": 0.337890625,
      "learning_rate": 1.2671023916975598e-05,
      "loss": 0.6681,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3475
    },
    {
      "epoch": 0.4726679358172423,
      "grad_norm": 0.3515625,
      "learning_rate": 1.2666448351190259e-05,
      "loss": 0.7482,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3476
    },
    {
      "epoch": 0.472803916236062,
      "grad_norm": 0.298828125,
      "learning_rate": 1.2661872184354862e-05,
      "loss": 0.6095,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3477
    },
    {
      "epoch": 0.4729398966548817,
      "grad_norm": 0.61328125,
      "learning_rate": 1.265729541750093e-05,
      "loss": 0.6854,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3478
    },
    {
      "epoch": 0.4730758770737014,
      "grad_norm": 0.369140625,
      "learning_rate": 1.2652718051660122e-05,
      "loss": 0.6364,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3479
    },
    {
      "epoch": 0.4732118574925211,
      "grad_norm": 0.81640625,
      "learning_rate": 1.2648140087864237e-05,
      "loss": 0.7674,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3480
    },
    {
      "epoch": 0.4733478379113408,
      "grad_norm": 0.4140625,
      "learning_rate": 1.2643561527145194e-05,
      "loss": 0.6349,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3481
    },
    {
      "epoch": 0.47348381833016046,
      "grad_norm": 0.53125,
      "learning_rate": 1.2638982370535064e-05,
      "loss": 0.6972,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3482
    },
    {
      "epoch": 0.47361979874898014,
      "grad_norm": 0.46875,
      "learning_rate": 1.2634402619066047e-05,
      "loss": 0.7992,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3483
    },
    {
      "epoch": 0.4737557791677998,
      "grad_norm": 0.28125,
      "learning_rate": 1.2629822273770467e-05,
      "loss": 0.6055,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3484
    },
    {
      "epoch": 0.47389175958661955,
      "grad_norm": 0.73828125,
      "learning_rate": 1.2625241335680796e-05,
      "loss": 0.9945,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3485
    },
    {
      "epoch": 0.47402774000543924,
      "grad_norm": 0.4296875,
      "learning_rate": 1.2620659805829631e-05,
      "loss": 0.6488,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3486
    },
    {
      "epoch": 0.4741637204242589,
      "grad_norm": 0.31640625,
      "learning_rate": 1.2616077685249705e-05,
      "loss": 0.5368,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3487
    },
    {
      "epoch": 0.4742997008430786,
      "grad_norm": 0.330078125,
      "learning_rate": 1.2611494974973884e-05,
      "loss": 0.6333,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3488
    },
    {
      "epoch": 0.4744356812618983,
      "grad_norm": 0.4921875,
      "learning_rate": 1.2606911676035165e-05,
      "loss": 0.7301,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3489
    },
    {
      "epoch": 0.47457166168071796,
      "grad_norm": 0.294921875,
      "learning_rate": 1.2602327789466686e-05,
      "loss": 0.5049,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3490
    },
    {
      "epoch": 0.47470764209953764,
      "grad_norm": 0.2578125,
      "learning_rate": 1.2597743316301705e-05,
      "loss": 0.3498,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3491
    },
    {
      "epoch": 0.4748436225183574,
      "grad_norm": 0.30078125,
      "learning_rate": 1.259315825757362e-05,
      "loss": 0.4194,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3492
    },
    {
      "epoch": 0.47497960293717706,
      "grad_norm": 0.54296875,
      "learning_rate": 1.2588572614315961e-05,
      "loss": 0.8433,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3493
    },
    {
      "epoch": 0.47511558335599674,
      "grad_norm": 0.59375,
      "learning_rate": 1.2583986387562388e-05,
      "loss": 0.5638,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3494
    },
    {
      "epoch": 0.4752515637748164,
      "grad_norm": 0.6875,
      "learning_rate": 1.2579399578346689e-05,
      "loss": 0.527,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3495
    },
    {
      "epoch": 0.4753875441936361,
      "grad_norm": 0.353515625,
      "learning_rate": 1.2574812187702789e-05,
      "loss": 0.5864,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3496
    },
    {
      "epoch": 0.4755235246124558,
      "grad_norm": 0.5,
      "learning_rate": 1.2570224216664744e-05,
      "loss": 0.6958,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3497
    },
    {
      "epoch": 0.4756595050312755,
      "grad_norm": 0.55859375,
      "learning_rate": 1.2565635666266736e-05,
      "loss": 0.6053,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3498
    },
    {
      "epoch": 0.4757954854500952,
      "grad_norm": 0.55078125,
      "learning_rate": 1.2561046537543078e-05,
      "loss": 0.7085,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3499
    },
    {
      "epoch": 0.4759314658689149,
      "grad_norm": 0.384765625,
      "learning_rate": 1.2556456831528223e-05,
      "loss": 0.7171,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3500
    },
    {
      "epoch": 0.47606744628773456,
      "grad_norm": 0.34765625,
      "learning_rate": 1.2551866549256739e-05,
      "loss": 0.6973,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3501
    },
    {
      "epoch": 0.47620342670655424,
      "grad_norm": 0.48828125,
      "learning_rate": 1.2547275691763339e-05,
      "loss": 0.6112,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3502
    },
    {
      "epoch": 0.4763394071253739,
      "grad_norm": 0.310546875,
      "learning_rate": 1.2542684260082853e-05,
      "loss": 0.3626,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3503
    },
    {
      "epoch": 0.47647538754419366,
      "grad_norm": 0.41015625,
      "learning_rate": 1.2538092255250249e-05,
      "loss": 0.7322,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3504
    },
    {
      "epoch": 0.47661136796301334,
      "grad_norm": 0.431640625,
      "learning_rate": 1.2533499678300618e-05,
      "loss": 0.6277,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3505
    },
    {
      "epoch": 0.476747348381833,
      "grad_norm": 0.92578125,
      "learning_rate": 1.2528906530269187e-05,
      "loss": 0.8543,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3506
    },
    {
      "epoch": 0.4768833288006527,
      "grad_norm": 0.4765625,
      "learning_rate": 1.2524312812191303e-05,
      "loss": 0.7212,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3507
    },
    {
      "epoch": 0.4770193092194724,
      "grad_norm": 0.45703125,
      "learning_rate": 1.2519718525102452e-05,
      "loss": 0.6921,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3508
    },
    {
      "epoch": 0.47715528963829207,
      "grad_norm": 0.78125,
      "learning_rate": 1.2515123670038236e-05,
      "loss": 0.5863,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3509
    },
    {
      "epoch": 0.47729127005711175,
      "grad_norm": 0.431640625,
      "learning_rate": 1.2510528248034395e-05,
      "loss": 0.7708,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3510
    },
    {
      "epoch": 0.4774272504759315,
      "grad_norm": 0.4375,
      "learning_rate": 1.2505932260126794e-05,
      "loss": 0.6893,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3511
    },
    {
      "epoch": 0.47756323089475117,
      "grad_norm": 0.443359375,
      "learning_rate": 1.2501335707351423e-05,
      "loss": 0.692,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3512
    },
    {
      "epoch": 0.47769921131357085,
      "grad_norm": 0.251953125,
      "learning_rate": 1.2496738590744402e-05,
      "loss": 0.4588,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3513
    },
    {
      "epoch": 0.47783519173239053,
      "grad_norm": 0.5,
      "learning_rate": 1.2492140911341979e-05,
      "loss": 0.8299,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3514
    },
    {
      "epoch": 0.4779711721512102,
      "grad_norm": 0.322265625,
      "learning_rate": 1.2487542670180523e-05,
      "loss": 0.5848,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3515
    },
    {
      "epoch": 0.4781071525700299,
      "grad_norm": 0.59765625,
      "learning_rate": 1.2482943868296537e-05,
      "loss": 0.8222,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3516
    },
    {
      "epoch": 0.47824313298884963,
      "grad_norm": 0.337890625,
      "learning_rate": 1.247834450672665e-05,
      "loss": 0.6502,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3517
    },
    {
      "epoch": 0.4783791134076693,
      "grad_norm": 0.55859375,
      "learning_rate": 1.2473744586507606e-05,
      "loss": 0.7021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3518
    },
    {
      "epoch": 0.478515093826489,
      "grad_norm": 0.474609375,
      "learning_rate": 1.2469144108676288e-05,
      "loss": 0.8353,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3519
    },
    {
      "epoch": 0.47865107424530867,
      "grad_norm": 0.455078125,
      "learning_rate": 1.2464543074269702e-05,
      "loss": 0.596,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3520
    },
    {
      "epoch": 0.47878705466412835,
      "grad_norm": 0.59375,
      "learning_rate": 1.2459941484324975e-05,
      "loss": 0.6224,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3521
    },
    {
      "epoch": 0.47892303508294803,
      "grad_norm": 0.46484375,
      "learning_rate": 1.2455339339879362e-05,
      "loss": 0.7783,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3522
    },
    {
      "epoch": 0.47905901550176777,
      "grad_norm": 0.54296875,
      "learning_rate": 1.2450736641970243e-05,
      "loss": 0.792,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3523
    },
    {
      "epoch": 0.47919499592058745,
      "grad_norm": 0.447265625,
      "learning_rate": 1.244613339163512e-05,
      "loss": 0.6134,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3524
    },
    {
      "epoch": 0.47933097633940713,
      "grad_norm": 0.60546875,
      "learning_rate": 1.2441529589911628e-05,
      "loss": 0.7316,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3525
    },
    {
      "epoch": 0.4794669567582268,
      "grad_norm": 0.359375,
      "learning_rate": 1.2436925237837512e-05,
      "loss": 0.6561,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3526
    },
    {
      "epoch": 0.4796029371770465,
      "grad_norm": 0.419921875,
      "learning_rate": 1.2432320336450656e-05,
      "loss": 0.6877,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3527
    },
    {
      "epoch": 0.4797389175958662,
      "grad_norm": 0.6015625,
      "learning_rate": 1.2427714886789057e-05,
      "loss": 0.8042,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3528
    },
    {
      "epoch": 0.47987489801468586,
      "grad_norm": 0.5,
      "learning_rate": 1.242310888989084e-05,
      "loss": 0.8115,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3529
    },
    {
      "epoch": 0.4800108784335056,
      "grad_norm": 1.0859375,
      "learning_rate": 1.2418502346794257e-05,
      "loss": 0.745,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3530
    },
    {
      "epoch": 0.4801468588523253,
      "grad_norm": 0.3984375,
      "learning_rate": 1.2413895258537676e-05,
      "loss": 0.5402,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3531
    },
    {
      "epoch": 0.48028283927114496,
      "grad_norm": 0.470703125,
      "learning_rate": 1.240928762615959e-05,
      "loss": 0.7262,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3532
    },
    {
      "epoch": 0.48041881968996464,
      "grad_norm": 0.328125,
      "learning_rate": 1.2404679450698616e-05,
      "loss": 0.3812,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3533
    },
    {
      "epoch": 0.4805548001087843,
      "grad_norm": 0.4296875,
      "learning_rate": 1.2400070733193494e-05,
      "loss": 0.7209,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3534
    },
    {
      "epoch": 0.480690780527604,
      "grad_norm": 0.34375,
      "learning_rate": 1.2395461474683086e-05,
      "loss": 0.6426,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3535
    },
    {
      "epoch": 0.48082676094642374,
      "grad_norm": 0.447265625,
      "learning_rate": 1.2390851676206376e-05,
      "loss": 0.6919,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3536
    },
    {
      "epoch": 0.4809627413652434,
      "grad_norm": 0.359375,
      "learning_rate": 1.2386241338802468e-05,
      "loss": 0.6789,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3537
    },
    {
      "epoch": 0.4810987217840631,
      "grad_norm": 0.400390625,
      "learning_rate": 1.2381630463510585e-05,
      "loss": 0.5814,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3538
    },
    {
      "epoch": 0.4812347022028828,
      "grad_norm": 0.58203125,
      "learning_rate": 1.237701905137008e-05,
      "loss": 0.789,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3539
    },
    {
      "epoch": 0.48137068262170246,
      "grad_norm": 0.294921875,
      "learning_rate": 1.2372407103420418e-05,
      "loss": 0.4854,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3540
    },
    {
      "epoch": 0.48150666304052214,
      "grad_norm": 0.435546875,
      "learning_rate": 1.236779462070119e-05,
      "loss": 0.5747,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3541
    },
    {
      "epoch": 0.4816426434593419,
      "grad_norm": 0.306640625,
      "learning_rate": 1.2363181604252104e-05,
      "loss": 0.6022,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3542
    },
    {
      "epoch": 0.48177862387816156,
      "grad_norm": 0.65625,
      "learning_rate": 1.2358568055112995e-05,
      "loss": 0.8739,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3543
    },
    {
      "epoch": 0.48191460429698124,
      "grad_norm": 0.5859375,
      "learning_rate": 1.2353953974323807e-05,
      "loss": 0.7518,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3544
    },
    {
      "epoch": 0.4820505847158009,
      "grad_norm": 0.314453125,
      "learning_rate": 1.2349339362924614e-05,
      "loss": 0.6022,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3545
    },
    {
      "epoch": 0.4821865651346206,
      "grad_norm": 0.2890625,
      "learning_rate": 1.2344724221955607e-05,
      "loss": 0.5241,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3546
    },
    {
      "epoch": 0.4823225455534403,
      "grad_norm": 0.3671875,
      "learning_rate": 1.2340108552457093e-05,
      "loss": 0.5562,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3547
    },
    {
      "epoch": 0.48245852597226,
      "grad_norm": 0.43359375,
      "learning_rate": 1.2335492355469501e-05,
      "loss": 0.6354,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3548
    },
    {
      "epoch": 0.4825945063910797,
      "grad_norm": 0.455078125,
      "learning_rate": 1.2330875632033377e-05,
      "loss": 0.8029,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3549
    },
    {
      "epoch": 0.4827304868098994,
      "grad_norm": 0.55859375,
      "learning_rate": 1.2326258383189388e-05,
      "loss": 0.5912,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3550
    },
    {
      "epoch": 0.48286646722871907,
      "grad_norm": 0.640625,
      "learning_rate": 1.232164060997832e-05,
      "loss": 0.6141,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3551
    },
    {
      "epoch": 0.48300244764753875,
      "grad_norm": 0.4375,
      "learning_rate": 1.2317022313441074e-05,
      "loss": 0.6768,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3552
    },
    {
      "epoch": 0.4831384280663584,
      "grad_norm": 0.609375,
      "learning_rate": 1.231240349461867e-05,
      "loss": 0.6825,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3553
    },
    {
      "epoch": 0.4832744084851781,
      "grad_norm": 0.375,
      "learning_rate": 1.2307784154552244e-05,
      "loss": 0.651,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3554
    },
    {
      "epoch": 0.48341038890399785,
      "grad_norm": 0.302734375,
      "learning_rate": 1.2303164294283059e-05,
      "loss": 0.6027,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3555
    },
    {
      "epoch": 0.4835463693228175,
      "grad_norm": 0.39453125,
      "learning_rate": 1.2298543914852482e-05,
      "loss": 0.6771,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3556
    },
    {
      "epoch": 0.4836823497416372,
      "grad_norm": 0.361328125,
      "learning_rate": 1.2293923017302004e-05,
      "loss": 0.8223,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3557
    },
    {
      "epoch": 0.4838183301604569,
      "grad_norm": 0.2734375,
      "learning_rate": 1.2289301602673231e-05,
      "loss": 0.4939,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3558
    },
    {
      "epoch": 0.48395431057927657,
      "grad_norm": 0.41015625,
      "learning_rate": 1.2284679672007889e-05,
      "loss": 0.6146,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3559
    },
    {
      "epoch": 0.48409029099809625,
      "grad_norm": 0.41796875,
      "learning_rate": 1.2280057226347814e-05,
      "loss": 0.8631,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3560
    },
    {
      "epoch": 0.484226271416916,
      "grad_norm": 0.44140625,
      "learning_rate": 1.2275434266734967e-05,
      "loss": 0.6432,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3561
    },
    {
      "epoch": 0.48436225183573567,
      "grad_norm": 0.546875,
      "learning_rate": 1.2270810794211415e-05,
      "loss": 0.7161,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3562
    },
    {
      "epoch": 0.48449823225455535,
      "grad_norm": 0.2421875,
      "learning_rate": 1.2266186809819347e-05,
      "loss": 0.5002,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3563
    },
    {
      "epoch": 0.48463421267337503,
      "grad_norm": 0.5546875,
      "learning_rate": 1.2261562314601068e-05,
      "loss": 0.7285,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3564
    },
    {
      "epoch": 0.4847701930921947,
      "grad_norm": 0.44921875,
      "learning_rate": 1.2256937309598991e-05,
      "loss": 0.5083,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3565
    },
    {
      "epoch": 0.4849061735110144,
      "grad_norm": 0.271484375,
      "learning_rate": 1.225231179585565e-05,
      "loss": 0.5011,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3566
    },
    {
      "epoch": 0.48504215392983413,
      "grad_norm": 0.345703125,
      "learning_rate": 1.2247685774413694e-05,
      "loss": 0.4548,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3567
    },
    {
      "epoch": 0.4851781343486538,
      "grad_norm": 0.66015625,
      "learning_rate": 1.2243059246315884e-05,
      "loss": 0.832,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3568
    },
    {
      "epoch": 0.4853141147674735,
      "grad_norm": 0.333984375,
      "learning_rate": 1.2238432212605097e-05,
      "loss": 0.7493,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3569
    },
    {
      "epoch": 0.4854500951862932,
      "grad_norm": 0.75,
      "learning_rate": 1.223380467432432e-05,
      "loss": 0.7575,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3570
    },
    {
      "epoch": 0.48558607560511285,
      "grad_norm": 0.345703125,
      "learning_rate": 1.2229176632516662e-05,
      "loss": 0.6037,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3571
    },
    {
      "epoch": 0.48572205602393254,
      "grad_norm": 0.40625,
      "learning_rate": 1.2224548088225335e-05,
      "loss": 0.7059,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3572
    },
    {
      "epoch": 0.4858580364427522,
      "grad_norm": 0.5625,
      "learning_rate": 1.2219919042493671e-05,
      "loss": 0.4671,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3573
    },
    {
      "epoch": 0.48599401686157195,
      "grad_norm": 0.33984375,
      "learning_rate": 1.2215289496365115e-05,
      "loss": 0.6392,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3574
    },
    {
      "epoch": 0.48612999728039163,
      "grad_norm": 0.3359375,
      "learning_rate": 1.2210659450883221e-05,
      "loss": 0.5879,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3575
    },
    {
      "epoch": 0.4862659776992113,
      "grad_norm": 0.26953125,
      "learning_rate": 1.220602890709166e-05,
      "loss": 0.5734,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3576
    },
    {
      "epoch": 0.486401958118031,
      "grad_norm": 0.3671875,
      "learning_rate": 1.2201397866034212e-05,
      "loss": 0.7441,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3577
    },
    {
      "epoch": 0.4865379385368507,
      "grad_norm": 0.244140625,
      "learning_rate": 1.219676632875477e-05,
      "loss": 0.515,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3578
    },
    {
      "epoch": 0.48667391895567036,
      "grad_norm": 0.375,
      "learning_rate": 1.219213429629734e-05,
      "loss": 0.6401,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3579
    },
    {
      "epoch": 0.4868098993744901,
      "grad_norm": 0.28125,
      "learning_rate": 1.2187501769706038e-05,
      "loss": 0.5785,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3580
    },
    {
      "epoch": 0.4869458797933098,
      "grad_norm": 0.33203125,
      "learning_rate": 1.2182868750025092e-05,
      "loss": 0.562,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3581
    },
    {
      "epoch": 0.48708186021212946,
      "grad_norm": 0.3359375,
      "learning_rate": 1.2178235238298844e-05,
      "loss": 0.5544,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3582
    },
    {
      "epoch": 0.48721784063094914,
      "grad_norm": 0.349609375,
      "learning_rate": 1.217360123557174e-05,
      "loss": 0.5788,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3583
    },
    {
      "epoch": 0.4873538210497688,
      "grad_norm": 0.357421875,
      "learning_rate": 1.2168966742888344e-05,
      "loss": 0.5365,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3584
    },
    {
      "epoch": 0.4874898014685885,
      "grad_norm": 0.61328125,
      "learning_rate": 1.2164331761293326e-05,
      "loss": 0.7075,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3585
    },
    {
      "epoch": 0.48762578188740824,
      "grad_norm": 0.51953125,
      "learning_rate": 1.2159696291831469e-05,
      "loss": 0.6419,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3586
    },
    {
      "epoch": 0.4877617623062279,
      "grad_norm": 0.357421875,
      "learning_rate": 1.215506033554766e-05,
      "loss": 0.6717,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3587
    },
    {
      "epoch": 0.4878977427250476,
      "grad_norm": 0.263671875,
      "learning_rate": 1.2150423893486907e-05,
      "loss": 0.5861,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3588
    },
    {
      "epoch": 0.4880337231438673,
      "grad_norm": 0.4453125,
      "learning_rate": 1.2145786966694315e-05,
      "loss": 0.6986,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3589
    },
    {
      "epoch": 0.48816970356268696,
      "grad_norm": 0.3125,
      "learning_rate": 1.2141149556215107e-05,
      "loss": 0.5869,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3590
    },
    {
      "epoch": 0.48830568398150664,
      "grad_norm": 0.365234375,
      "learning_rate": 1.213651166309461e-05,
      "loss": 0.5343,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3591
    },
    {
      "epoch": 0.4884416644003263,
      "grad_norm": 0.34765625,
      "learning_rate": 1.2131873288378267e-05,
      "loss": 0.7316,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3592
    },
    {
      "epoch": 0.48857764481914606,
      "grad_norm": 0.359375,
      "learning_rate": 1.2127234433111616e-05,
      "loss": 0.6969,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3593
    },
    {
      "epoch": 0.48871362523796574,
      "grad_norm": 0.2470703125,
      "learning_rate": 1.2122595098340318e-05,
      "loss": 0.3893,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3594
    },
    {
      "epoch": 0.4888496056567854,
      "grad_norm": 0.5,
      "learning_rate": 1.2117955285110133e-05,
      "loss": 0.7992,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3595
    },
    {
      "epoch": 0.4889855860756051,
      "grad_norm": 0.294921875,
      "learning_rate": 1.211331499446693e-05,
      "loss": 0.3494,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3596
    },
    {
      "epoch": 0.4891215664944248,
      "grad_norm": 0.279296875,
      "learning_rate": 1.2108674227456693e-05,
      "loss": 0.4673,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3597
    },
    {
      "epoch": 0.48925754691324447,
      "grad_norm": 0.25390625,
      "learning_rate": 1.2104032985125505e-05,
      "loss": 0.5444,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3598
    },
    {
      "epoch": 0.4893935273320642,
      "grad_norm": 2.71875,
      "learning_rate": 1.2099391268519556e-05,
      "loss": 0.9268,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3599
    },
    {
      "epoch": 0.4895295077508839,
      "grad_norm": 0.380859375,
      "learning_rate": 1.2094749078685148e-05,
      "loss": 0.6509,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3600
    },
    {
      "epoch": 0.48966548816970357,
      "grad_norm": 0.3125,
      "learning_rate": 1.2090106416668688e-05,
      "loss": 0.6545,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3601
    },
    {
      "epoch": 0.48980146858852325,
      "grad_norm": 0.640625,
      "learning_rate": 1.2085463283516688e-05,
      "loss": 0.7209,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3602
    },
    {
      "epoch": 0.48993744900734293,
      "grad_norm": 0.31640625,
      "learning_rate": 1.2080819680275767e-05,
      "loss": 0.6286,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3603
    },
    {
      "epoch": 0.4900734294261626,
      "grad_norm": 0.1796875,
      "learning_rate": 1.2076175607992648e-05,
      "loss": 0.2826,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3604
    },
    {
      "epoch": 0.49020940984498235,
      "grad_norm": 0.53515625,
      "learning_rate": 1.2071531067714164e-05,
      "loss": 0.6494,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3605
    },
    {
      "epoch": 0.49034539026380203,
      "grad_norm": 0.345703125,
      "learning_rate": 1.2066886060487251e-05,
      "loss": 0.6222,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3606
    },
    {
      "epoch": 0.4904813706826217,
      "grad_norm": 0.40234375,
      "learning_rate": 1.2062240587358952e-05,
      "loss": 0.799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3607
    },
    {
      "epoch": 0.4906173511014414,
      "grad_norm": 0.2578125,
      "learning_rate": 1.2057594649376408e-05,
      "loss": 0.5256,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3608
    },
    {
      "epoch": 0.49075333152026107,
      "grad_norm": 0.30859375,
      "learning_rate": 1.2052948247586872e-05,
      "loss": 0.4849,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3609
    },
    {
      "epoch": 0.49088931193908075,
      "grad_norm": 0.412109375,
      "learning_rate": 1.2048301383037706e-05,
      "loss": 0.7384,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3610
    },
    {
      "epoch": 0.4910252923579005,
      "grad_norm": 0.61328125,
      "learning_rate": 1.2043654056776363e-05,
      "loss": 0.7855,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3611
    },
    {
      "epoch": 0.49116127277672017,
      "grad_norm": 0.333984375,
      "learning_rate": 1.2039006269850413e-05,
      "loss": 0.5698,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3612
    },
    {
      "epoch": 0.49129725319553985,
      "grad_norm": 0.453125,
      "learning_rate": 1.2034358023307517e-05,
      "loss": 0.7458,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3613
    },
    {
      "epoch": 0.49143323361435953,
      "grad_norm": 0.458984375,
      "learning_rate": 1.2029709318195452e-05,
      "loss": 0.7856,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3614
    },
    {
      "epoch": 0.4915692140331792,
      "grad_norm": 0.52734375,
      "learning_rate": 1.2025060155562092e-05,
      "loss": 0.6027,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3615
    },
    {
      "epoch": 0.4917051944519989,
      "grad_norm": 0.6875,
      "learning_rate": 1.2020410536455413e-05,
      "loss": 0.7495,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3616
    },
    {
      "epoch": 0.4918411748708186,
      "grad_norm": 0.59375,
      "learning_rate": 1.2015760461923499e-05,
      "loss": 0.6714,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3617
    },
    {
      "epoch": 0.4919771552896383,
      "grad_norm": 0.7109375,
      "learning_rate": 1.201110993301453e-05,
      "loss": 0.5073,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3618
    },
    {
      "epoch": 0.492113135708458,
      "grad_norm": 0.484375,
      "learning_rate": 1.2006458950776797e-05,
      "loss": 0.5507,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3619
    },
    {
      "epoch": 0.4922491161272777,
      "grad_norm": 0.58203125,
      "learning_rate": 1.2001807516258684e-05,
      "loss": 0.4731,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3620
    },
    {
      "epoch": 0.49238509654609736,
      "grad_norm": 0.34765625,
      "learning_rate": 1.199715563050868e-05,
      "loss": 0.626,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3621
    },
    {
      "epoch": 0.49252107696491704,
      "grad_norm": 0.384765625,
      "learning_rate": 1.1992503294575385e-05,
      "loss": 0.7297,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3622
    },
    {
      "epoch": 0.4926570573837367,
      "grad_norm": 0.1826171875,
      "learning_rate": 1.1987850509507481e-05,
      "loss": 0.258,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3623
    },
    {
      "epoch": 0.49279303780255646,
      "grad_norm": 0.341796875,
      "learning_rate": 1.1983197276353773e-05,
      "loss": 0.4006,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3624
    },
    {
      "epoch": 0.49292901822137614,
      "grad_norm": 0.302734375,
      "learning_rate": 1.197854359616315e-05,
      "loss": 0.6068,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3625
    },
    {
      "epoch": 0.4930649986401958,
      "grad_norm": 0.54296875,
      "learning_rate": 1.197388946998461e-05,
      "loss": 0.6696,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3626
    },
    {
      "epoch": 0.4932009790590155,
      "grad_norm": 0.330078125,
      "learning_rate": 1.1969234898867252e-05,
      "loss": 0.5669,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3627
    },
    {
      "epoch": 0.4933369594778352,
      "grad_norm": 0.427734375,
      "learning_rate": 1.196457988386027e-05,
      "loss": 0.9554,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3628
    },
    {
      "epoch": 0.49347293989665486,
      "grad_norm": 0.439453125,
      "learning_rate": 1.1959924426012965e-05,
      "loss": 0.647,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3629
    },
    {
      "epoch": 0.4936089203154746,
      "grad_norm": 0.3515625,
      "learning_rate": 1.1955268526374733e-05,
      "loss": 0.6997,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3630
    },
    {
      "epoch": 0.4937449007342943,
      "grad_norm": 0.267578125,
      "learning_rate": 1.1950612185995067e-05,
      "loss": 0.4819,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3631
    },
    {
      "epoch": 0.49388088115311396,
      "grad_norm": 0.5390625,
      "learning_rate": 1.1945955405923568e-05,
      "loss": 0.5853,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3632
    },
    {
      "epoch": 0.49401686157193364,
      "grad_norm": 0.453125,
      "learning_rate": 1.194129818720993e-05,
      "loss": 0.6206,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3633
    },
    {
      "epoch": 0.4941528419907533,
      "grad_norm": 0.1845703125,
      "learning_rate": 1.1936640530903942e-05,
      "loss": 0.3558,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3634
    },
    {
      "epoch": 0.494288822409573,
      "grad_norm": 0.392578125,
      "learning_rate": 1.1931982438055506e-05,
      "loss": 0.66,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3635
    },
    {
      "epoch": 0.4944248028283927,
      "grad_norm": 0.51953125,
      "learning_rate": 1.1927323909714603e-05,
      "loss": 0.585,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3636
    },
    {
      "epoch": 0.4945607832472124,
      "grad_norm": 0.353515625,
      "learning_rate": 1.192266494693133e-05,
      "loss": 0.7534,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3637
    },
    {
      "epoch": 0.4946967636660321,
      "grad_norm": 0.6484375,
      "learning_rate": 1.1918005550755871e-05,
      "loss": 0.6014,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3638
    },
    {
      "epoch": 0.4948327440848518,
      "grad_norm": 0.318359375,
      "learning_rate": 1.1913345722238509e-05,
      "loss": 0.6107,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3639
    },
    {
      "epoch": 0.49496872450367146,
      "grad_norm": 0.34765625,
      "learning_rate": 1.190868546242963e-05,
      "loss": 0.5951,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3640
    },
    {
      "epoch": 0.49510470492249115,
      "grad_norm": 0.6953125,
      "learning_rate": 1.1904024772379713e-05,
      "loss": 0.9429,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3641
    },
    {
      "epoch": 0.4952406853413108,
      "grad_norm": 3.921875,
      "learning_rate": 1.1899363653139334e-05,
      "loss": 0.6525,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3642
    },
    {
      "epoch": 0.49537666576013056,
      "grad_norm": 0.296875,
      "learning_rate": 1.1894702105759164e-05,
      "loss": 0.5895,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3643
    },
    {
      "epoch": 0.49551264617895024,
      "grad_norm": 0.478515625,
      "learning_rate": 1.1890040131289975e-05,
      "loss": 0.5579,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3644
    },
    {
      "epoch": 0.4956486265977699,
      "grad_norm": 0.71875,
      "learning_rate": 1.1885377730782632e-05,
      "loss": 0.4544,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3645
    },
    {
      "epoch": 0.4957846070165896,
      "grad_norm": 0.390625,
      "learning_rate": 1.1880714905288102e-05,
      "loss": 0.6772,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3646
    },
    {
      "epoch": 0.4959205874354093,
      "grad_norm": 0.388671875,
      "learning_rate": 1.1876051655857434e-05,
      "loss": 0.5907,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3647
    },
    {
      "epoch": 0.49605656785422897,
      "grad_norm": 0.58203125,
      "learning_rate": 1.1871387983541789e-05,
      "loss": 0.4346,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3648
    },
    {
      "epoch": 0.4961925482730487,
      "grad_norm": 0.3515625,
      "learning_rate": 1.186672388939241e-05,
      "loss": 0.6649,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3649
    },
    {
      "epoch": 0.4963285286918684,
      "grad_norm": 0.4453125,
      "learning_rate": 1.1862059374460644e-05,
      "loss": 0.3963,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3650
    },
    {
      "epoch": 0.49646450911068807,
      "grad_norm": 0.412109375,
      "learning_rate": 1.185739443979793e-05,
      "loss": 0.7865,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3651
    },
    {
      "epoch": 0.49660048952950775,
      "grad_norm": 0.41015625,
      "learning_rate": 1.1852729086455801e-05,
      "loss": 0.6331,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3652
    },
    {
      "epoch": 0.49673646994832743,
      "grad_norm": 0.25,
      "learning_rate": 1.1848063315485884e-05,
      "loss": 0.4756,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3653
    },
    {
      "epoch": 0.4968724503671471,
      "grad_norm": 0.8359375,
      "learning_rate": 1.1843397127939901e-05,
      "loss": 0.7158,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3654
    },
    {
      "epoch": 0.49700843078596685,
      "grad_norm": 0.416015625,
      "learning_rate": 1.1838730524869667e-05,
      "loss": 0.71,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3655
    },
    {
      "epoch": 0.49714441120478653,
      "grad_norm": 0.37109375,
      "learning_rate": 1.1834063507327093e-05,
      "loss": 0.6234,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3656
    },
    {
      "epoch": 0.4972803916236062,
      "grad_norm": 0.423828125,
      "learning_rate": 1.182939607636418e-05,
      "loss": 0.8034,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3657
    },
    {
      "epoch": 0.4974163720424259,
      "grad_norm": 0.330078125,
      "learning_rate": 1.1824728233033027e-05,
      "loss": 0.6196,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3658
    },
    {
      "epoch": 0.4975523524612456,
      "grad_norm": 0.4921875,
      "learning_rate": 1.1820059978385821e-05,
      "loss": 0.7686,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3659
    },
    {
      "epoch": 0.49768833288006525,
      "grad_norm": 0.36328125,
      "learning_rate": 1.1815391313474845e-05,
      "loss": 0.6455,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3660
    },
    {
      "epoch": 0.49782431329888494,
      "grad_norm": 0.330078125,
      "learning_rate": 1.1810722239352468e-05,
      "loss": 0.5964,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3661
    },
    {
      "epoch": 0.49796029371770467,
      "grad_norm": 0.361328125,
      "learning_rate": 1.1806052757071163e-05,
      "loss": 0.7799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3662
    },
    {
      "epoch": 0.49809627413652435,
      "grad_norm": 0.4609375,
      "learning_rate": 1.1801382867683484e-05,
      "loss": 0.7588,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3663
    },
    {
      "epoch": 0.49823225455534403,
      "grad_norm": 0.375,
      "learning_rate": 1.1796712572242082e-05,
      "loss": 0.6896,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3664
    },
    {
      "epoch": 0.4983682349741637,
      "grad_norm": 0.396484375,
      "learning_rate": 1.1792041871799701e-05,
      "loss": 0.5539,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3665
    },
    {
      "epoch": 0.4985042153929834,
      "grad_norm": 0.2734375,
      "learning_rate": 1.1787370767409171e-05,
      "loss": 0.4163,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3666
    },
    {
      "epoch": 0.4986401958118031,
      "grad_norm": 0.28515625,
      "learning_rate": 1.178269926012342e-05,
      "loss": 0.6519,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3667
    },
    {
      "epoch": 0.4987761762306228,
      "grad_norm": 0.29296875,
      "learning_rate": 1.177802735099546e-05,
      "loss": 0.5021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3668
    },
    {
      "epoch": 0.4989121566494425,
      "grad_norm": 0.625,
      "learning_rate": 1.1773355041078394e-05,
      "loss": 0.6131,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3669
    },
    {
      "epoch": 0.4990481370682622,
      "grad_norm": 0.609375,
      "learning_rate": 1.176868233142542e-05,
      "loss": 0.9064,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3670
    },
    {
      "epoch": 0.49918411748708186,
      "grad_norm": 0.470703125,
      "learning_rate": 1.1764009223089827e-05,
      "loss": 0.7738,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3671
    },
    {
      "epoch": 0.49932009790590154,
      "grad_norm": 0.3203125,
      "learning_rate": 1.1759335717124988e-05,
      "loss": 0.5897,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3672
    },
    {
      "epoch": 0.4994560783247212,
      "grad_norm": 0.6171875,
      "learning_rate": 1.175466181458437e-05,
      "loss": 0.6401,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3673
    },
    {
      "epoch": 0.49959205874354096,
      "grad_norm": 0.3828125,
      "learning_rate": 1.1749987516521523e-05,
      "loss": 0.6619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3674
    },
    {
      "epoch": 0.49972803916236064,
      "grad_norm": 0.50390625,
      "learning_rate": 1.1745312823990099e-05,
      "loss": 0.863,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3675
    },
    {
      "epoch": 0.4998640195811803,
      "grad_norm": 0.3359375,
      "learning_rate": 1.1740637738043822e-05,
      "loss": 0.6934,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3676
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.68359375,
      "learning_rate": 1.173596225973652e-05,
      "loss": 0.771,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3677
    },
    {
      "epoch": 0.5001359804188197,
      "grad_norm": 0.38671875,
      "learning_rate": 1.1731286390122099e-05,
      "loss": 0.5417,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3678
    },
    {
      "epoch": 0.5002719608376394,
      "grad_norm": 0.94140625,
      "learning_rate": 1.172661013025456e-05,
      "loss": 0.7946,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3679
    },
    {
      "epoch": 0.500407941256459,
      "grad_norm": 0.56640625,
      "learning_rate": 1.1721933481187988e-05,
      "loss": 0.7559,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3680
    },
    {
      "epoch": 0.5005439216752787,
      "grad_norm": 0.384765625,
      "learning_rate": 1.1717256443976555e-05,
      "loss": 0.5374,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3681
    },
    {
      "epoch": 0.5006799020940984,
      "grad_norm": 0.65234375,
      "learning_rate": 1.1712579019674527e-05,
      "loss": 0.5111,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3682
    },
    {
      "epoch": 0.5008158825129181,
      "grad_norm": 0.34375,
      "learning_rate": 1.1707901209336247e-05,
      "loss": 0.5981,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3683
    },
    {
      "epoch": 0.5009518629317379,
      "grad_norm": 0.42578125,
      "learning_rate": 1.1703223014016154e-05,
      "loss": 0.6361,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3684
    },
    {
      "epoch": 0.5010878433505576,
      "grad_norm": 0.3984375,
      "learning_rate": 1.1698544434768773e-05,
      "loss": 0.7518,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3685
    },
    {
      "epoch": 0.5012238237693772,
      "grad_norm": 0.34375,
      "learning_rate": 1.1693865472648707e-05,
      "loss": 0.7311,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3686
    },
    {
      "epoch": 0.5013598041881969,
      "grad_norm": 0.40234375,
      "learning_rate": 1.1689186128710654e-05,
      "loss": 0.7487,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3687
    },
    {
      "epoch": 0.5014957846070166,
      "grad_norm": 0.341796875,
      "learning_rate": 1.1684506404009397e-05,
      "loss": 0.5781,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3688
    },
    {
      "epoch": 0.5016317650258363,
      "grad_norm": 0.3359375,
      "learning_rate": 1.1679826299599801e-05,
      "loss": 0.5524,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3689
    },
    {
      "epoch": 0.501767745444656,
      "grad_norm": 0.357421875,
      "learning_rate": 1.167514581653682e-05,
      "loss": 0.542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3690
    },
    {
      "epoch": 0.5019037258634756,
      "grad_norm": 0.30078125,
      "learning_rate": 1.167046495587549e-05,
      "loss": 0.5382,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3691
    },
    {
      "epoch": 0.5020397062822953,
      "grad_norm": 0.7109375,
      "learning_rate": 1.1665783718670936e-05,
      "loss": 0.7295,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3692
    },
    {
      "epoch": 0.502175686701115,
      "grad_norm": 0.73046875,
      "learning_rate": 1.1661102105978367e-05,
      "loss": 0.9139,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3693
    },
    {
      "epoch": 0.5023116671199347,
      "grad_norm": 0.447265625,
      "learning_rate": 1.1656420118853073e-05,
      "loss": 0.7463,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3694
    },
    {
      "epoch": 0.5024476475387544,
      "grad_norm": 0.48046875,
      "learning_rate": 1.1651737758350432e-05,
      "loss": 0.6141,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3695
    },
    {
      "epoch": 0.5025836279575742,
      "grad_norm": 0.6015625,
      "learning_rate": 1.1647055025525908e-05,
      "loss": 0.8273,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3696
    },
    {
      "epoch": 0.5027196083763938,
      "grad_norm": 0.5859375,
      "learning_rate": 1.1642371921435045e-05,
      "loss": 0.6346,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3697
    },
    {
      "epoch": 0.5028555887952135,
      "grad_norm": 0.326171875,
      "learning_rate": 1.1637688447133468e-05,
      "loss": 0.5643,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3698
    },
    {
      "epoch": 0.5029915692140332,
      "grad_norm": 0.2890625,
      "learning_rate": 1.1633004603676895e-05,
      "loss": 0.5369,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3699
    },
    {
      "epoch": 0.5031275496328529,
      "grad_norm": 1.65625,
      "learning_rate": 1.1628320392121118e-05,
      "loss": 0.8101,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3700
    },
    {
      "epoch": 0.5032635300516726,
      "grad_norm": 0.59375,
      "learning_rate": 1.1623635813522019e-05,
      "loss": 0.7257,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3701
    },
    {
      "epoch": 0.5033995104704923,
      "grad_norm": 0.267578125,
      "learning_rate": 1.1618950868935554e-05,
      "loss": 0.4998,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3702
    },
    {
      "epoch": 0.5035354908893119,
      "grad_norm": 0.384765625,
      "learning_rate": 1.161426555941777e-05,
      "loss": 0.6232,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3703
    },
    {
      "epoch": 0.5036714713081316,
      "grad_norm": 0.36328125,
      "learning_rate": 1.1609579886024793e-05,
      "loss": 0.6374,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3704
    },
    {
      "epoch": 0.5038074517269513,
      "grad_norm": 0.2734375,
      "learning_rate": 1.1604893849812832e-05,
      "loss": 0.4243,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3705
    },
    {
      "epoch": 0.503943432145771,
      "grad_norm": 0.2275390625,
      "learning_rate": 1.1600207451838174e-05,
      "loss": 0.4012,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3706
    },
    {
      "epoch": 0.5040794125645907,
      "grad_norm": 0.35546875,
      "learning_rate": 1.1595520693157194e-05,
      "loss": 0.6738,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3707
    },
    {
      "epoch": 0.5042153929834103,
      "grad_norm": 0.412109375,
      "learning_rate": 1.1590833574826341e-05,
      "loss": 0.6529,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3708
    },
    {
      "epoch": 0.5043513734022301,
      "grad_norm": 0.375,
      "learning_rate": 1.1586146097902153e-05,
      "loss": 0.6027,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3709
    },
    {
      "epoch": 0.5044873538210498,
      "grad_norm": 0.267578125,
      "learning_rate": 1.158145826344124e-05,
      "loss": 0.4334,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3710
    },
    {
      "epoch": 0.5046233342398695,
      "grad_norm": 0.322265625,
      "learning_rate": 1.1576770072500301e-05,
      "loss": 0.5334,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3711
    },
    {
      "epoch": 0.5047593146586892,
      "grad_norm": 0.69921875,
      "learning_rate": 1.1572081526136108e-05,
      "loss": 0.7253,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3712
    },
    {
      "epoch": 0.5048952950775089,
      "grad_norm": 0.298828125,
      "learning_rate": 1.156739262540552e-05,
      "loss": 0.5163,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3713
    },
    {
      "epoch": 0.5050312754963285,
      "grad_norm": 0.291015625,
      "learning_rate": 1.1562703371365472e-05,
      "loss": 0.5236,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3714
    },
    {
      "epoch": 0.5051672559151482,
      "grad_norm": 0.37109375,
      "learning_rate": 1.1558013765072974e-05,
      "loss": 0.7214,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3715
    },
    {
      "epoch": 0.5053032363339679,
      "grad_norm": 0.390625,
      "learning_rate": 1.1553323807585128e-05,
      "loss": 0.6831,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3716
    },
    {
      "epoch": 0.5054392167527876,
      "grad_norm": 0.87890625,
      "learning_rate": 1.1548633499959105e-05,
      "loss": 0.7275,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3717
    },
    {
      "epoch": 0.5055751971716073,
      "grad_norm": 1.1953125,
      "learning_rate": 1.1543942843252158e-05,
      "loss": 0.5711,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3718
    },
    {
      "epoch": 0.5057111775904269,
      "grad_norm": 0.4375,
      "learning_rate": 1.1539251838521617e-05,
      "loss": 0.7262,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3719
    },
    {
      "epoch": 0.5058471580092466,
      "grad_norm": 0.375,
      "learning_rate": 1.1534560486824892e-05,
      "loss": 0.609,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3720
    },
    {
      "epoch": 0.5059831384280663,
      "grad_norm": 0.39453125,
      "learning_rate": 1.1529868789219474e-05,
      "loss": 0.6802,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3721
    },
    {
      "epoch": 0.5061191188468861,
      "grad_norm": 0.435546875,
      "learning_rate": 1.1525176746762922e-05,
      "loss": 0.7549,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3722
    },
    {
      "epoch": 0.5062550992657058,
      "grad_norm": 0.7421875,
      "learning_rate": 1.152048436051289e-05,
      "loss": 0.6824,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3723
    },
    {
      "epoch": 0.5063910796845255,
      "grad_norm": 0.326171875,
      "learning_rate": 1.1515791631527089e-05,
      "loss": 0.5762,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3724
    },
    {
      "epoch": 0.5065270601033451,
      "grad_norm": 0.69921875,
      "learning_rate": 1.1511098560863325e-05,
      "loss": 0.7282,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3725
    },
    {
      "epoch": 0.5066630405221648,
      "grad_norm": 0.310546875,
      "learning_rate": 1.150640514957947e-05,
      "loss": 0.6559,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3726
    },
    {
      "epoch": 0.5067990209409845,
      "grad_norm": 0.392578125,
      "learning_rate": 1.1501711398733474e-05,
      "loss": 0.6496,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3727
    },
    {
      "epoch": 0.5069350013598042,
      "grad_norm": 0.58984375,
      "learning_rate": 1.1497017309383372e-05,
      "loss": 0.5516,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3728
    },
    {
      "epoch": 0.5070709817786239,
      "grad_norm": 0.400390625,
      "learning_rate": 1.149232288258726e-05,
      "loss": 0.4657,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3729
    },
    {
      "epoch": 0.5072069621974435,
      "grad_norm": 1.34375,
      "learning_rate": 1.148762811940333e-05,
      "loss": 0.779,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3730
    },
    {
      "epoch": 0.5073429426162632,
      "grad_norm": 0.578125,
      "learning_rate": 1.1482933020889835e-05,
      "loss": 0.5806,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3731
    },
    {
      "epoch": 0.5074789230350829,
      "grad_norm": 0.2373046875,
      "learning_rate": 1.1478237588105105e-05,
      "loss": 0.3823,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3732
    },
    {
      "epoch": 0.5076149034539026,
      "grad_norm": 0.3828125,
      "learning_rate": 1.147354182210755e-05,
      "loss": 0.5736,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3733
    },
    {
      "epoch": 0.5077508838727224,
      "grad_norm": 0.6484375,
      "learning_rate": 1.1468845723955652e-05,
      "loss": 0.5877,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3734
    },
    {
      "epoch": 0.5078868642915421,
      "grad_norm": 0.396484375,
      "learning_rate": 1.1464149294707972e-05,
      "loss": 0.6899,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3735
    },
    {
      "epoch": 0.5080228447103617,
      "grad_norm": 0.2734375,
      "learning_rate": 1.1459452535423137e-05,
      "loss": 0.5475,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3736
    },
    {
      "epoch": 0.5081588251291814,
      "grad_norm": 0.50390625,
      "learning_rate": 1.145475544715986e-05,
      "loss": 0.7004,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3737
    },
    {
      "epoch": 0.5082948055480011,
      "grad_norm": 0.357421875,
      "learning_rate": 1.1450058030976916e-05,
      "loss": 0.6139,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3738
    },
    {
      "epoch": 0.5084307859668208,
      "grad_norm": 0.578125,
      "learning_rate": 1.1445360287933165e-05,
      "loss": 0.5726,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3739
    },
    {
      "epoch": 0.5085667663856405,
      "grad_norm": 0.33984375,
      "learning_rate": 1.1440662219087535e-05,
      "loss": 0.7021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3740
    },
    {
      "epoch": 0.5087027468044601,
      "grad_norm": 0.5546875,
      "learning_rate": 1.1435963825499026e-05,
      "loss": 0.5425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3741
    },
    {
      "epoch": 0.5088387272232798,
      "grad_norm": 0.21875,
      "learning_rate": 1.1431265108226715e-05,
      "loss": 0.3704,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3742
    },
    {
      "epoch": 0.5089747076420995,
      "grad_norm": 0.353515625,
      "learning_rate": 1.142656606832975e-05,
      "loss": 0.6177,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3743
    },
    {
      "epoch": 0.5091106880609192,
      "grad_norm": 0.4296875,
      "learning_rate": 1.142186670686735e-05,
      "loss": 0.7082,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3744
    },
    {
      "epoch": 0.5092466684797389,
      "grad_norm": 0.470703125,
      "learning_rate": 1.1417167024898812e-05,
      "loss": 0.7269,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3745
    },
    {
      "epoch": 0.5093826488985586,
      "grad_norm": 0.416015625,
      "learning_rate": 1.14124670234835e-05,
      "loss": 0.7173,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3746
    },
    {
      "epoch": 0.5095186293173783,
      "grad_norm": 0.578125,
      "learning_rate": 1.1407766703680852e-05,
      "loss": 0.6255,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3747
    },
    {
      "epoch": 0.509654609736198,
      "grad_norm": 0.341796875,
      "learning_rate": 1.1403066066550377e-05,
      "loss": 0.5801,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3748
    },
    {
      "epoch": 0.5097905901550177,
      "grad_norm": 0.462890625,
      "learning_rate": 1.1398365113151657e-05,
      "loss": 0.6353,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3749
    },
    {
      "epoch": 0.5099265705738374,
      "grad_norm": 0.60546875,
      "learning_rate": 1.1393663844544345e-05,
      "loss": 0.5837,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3750
    },
    {
      "epoch": 0.5100625509926571,
      "grad_norm": 0.259765625,
      "learning_rate": 1.1388962261788163e-05,
      "loss": 0.4305,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3751
    },
    {
      "epoch": 0.5101985314114768,
      "grad_norm": 0.44140625,
      "learning_rate": 1.1384260365942905e-05,
      "loss": 0.8978,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3752
    },
    {
      "epoch": 0.5103345118302964,
      "grad_norm": 0.388671875,
      "learning_rate": 1.1379558158068436e-05,
      "loss": 0.7295,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3753
    },
    {
      "epoch": 0.5104704922491161,
      "grad_norm": 0.4296875,
      "learning_rate": 1.1374855639224691e-05,
      "loss": 0.7607,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3754
    },
    {
      "epoch": 0.5106064726679358,
      "grad_norm": 0.41015625,
      "learning_rate": 1.1370152810471677e-05,
      "loss": 0.7866,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3755
    },
    {
      "epoch": 0.5107424530867555,
      "grad_norm": 0.3828125,
      "learning_rate": 1.1365449672869469e-05,
      "loss": 0.6625,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3756
    },
    {
      "epoch": 0.5108784335055752,
      "grad_norm": 0.2373046875,
      "learning_rate": 1.136074622747821e-05,
      "loss": 0.4714,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3757
    },
    {
      "epoch": 0.5110144139243948,
      "grad_norm": 0.279296875,
      "learning_rate": 1.1356042475358113e-05,
      "loss": 0.5765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3758
    },
    {
      "epoch": 0.5111503943432146,
      "grad_norm": 0.310546875,
      "learning_rate": 1.1351338417569467e-05,
      "loss": 0.6092,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3759
    },
    {
      "epoch": 0.5112863747620343,
      "grad_norm": 0.318359375,
      "learning_rate": 1.1346634055172616e-05,
      "loss": 0.5627,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3760
    },
    {
      "epoch": 0.511422355180854,
      "grad_norm": 0.240234375,
      "learning_rate": 1.1341929389227985e-05,
      "loss": 0.3564,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3761
    },
    {
      "epoch": 0.5115583355996737,
      "grad_norm": 3.390625,
      "learning_rate": 1.1337224420796064e-05,
      "loss": 0.734,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3762
    },
    {
      "epoch": 0.5116943160184934,
      "grad_norm": 0.365234375,
      "learning_rate": 1.1332519150937411e-05,
      "loss": 0.6464,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3763
    },
    {
      "epoch": 0.511830296437313,
      "grad_norm": 0.380859375,
      "learning_rate": 1.132781358071265e-05,
      "loss": 0.7054,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3764
    },
    {
      "epoch": 0.5119662768561327,
      "grad_norm": 0.341796875,
      "learning_rate": 1.1323107711182474e-05,
      "loss": 0.6548,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3765
    },
    {
      "epoch": 0.5121022572749524,
      "grad_norm": 0.359375,
      "learning_rate": 1.131840154340764e-05,
      "loss": 0.7119,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3766
    },
    {
      "epoch": 0.5122382376937721,
      "grad_norm": 0.275390625,
      "learning_rate": 1.131369507844898e-05,
      "loss": 0.5706,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3767
    },
    {
      "epoch": 0.5123742181125918,
      "grad_norm": 0.287109375,
      "learning_rate": 1.130898831736739e-05,
      "loss": 0.5397,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3768
    },
    {
      "epoch": 0.5125101985314114,
      "grad_norm": 0.37109375,
      "learning_rate": 1.130428126122383e-05,
      "loss": 0.5381,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3769
    },
    {
      "epoch": 0.5126461789502311,
      "grad_norm": 0.69140625,
      "learning_rate": 1.1299573911079326e-05,
      "loss": 0.7012,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3770
    },
    {
      "epoch": 0.5127821593690508,
      "grad_norm": 0.50390625,
      "learning_rate": 1.1294866267994975e-05,
      "loss": 0.7144,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3771
    },
    {
      "epoch": 0.5129181397878706,
      "grad_norm": 0.33203125,
      "learning_rate": 1.1290158333031936e-05,
      "loss": 0.6217,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3772
    },
    {
      "epoch": 0.5130541202066903,
      "grad_norm": 0.2412109375,
      "learning_rate": 1.1285450107251434e-05,
      "loss": 0.5339,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3773
    },
    {
      "epoch": 0.51319010062551,
      "grad_norm": 0.494140625,
      "learning_rate": 1.1280741591714761e-05,
      "loss": 0.8332,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3774
    },
    {
      "epoch": 0.5133260810443296,
      "grad_norm": 0.32421875,
      "learning_rate": 1.1276032787483278e-05,
      "loss": 0.6408,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3775
    },
    {
      "epoch": 0.5134620614631493,
      "grad_norm": 2.6875,
      "learning_rate": 1.12713236956184e-05,
      "loss": 0.5524,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3776
    },
    {
      "epoch": 0.513598041881969,
      "grad_norm": 0.330078125,
      "learning_rate": 1.1266614317181619e-05,
      "loss": 0.6522,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3777
    },
    {
      "epoch": 0.5137340223007887,
      "grad_norm": 0.32421875,
      "learning_rate": 1.1261904653234486e-05,
      "loss": 0.5527,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3778
    },
    {
      "epoch": 0.5138700027196084,
      "grad_norm": 0.54296875,
      "learning_rate": 1.1257194704838616e-05,
      "loss": 0.618,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3779
    },
    {
      "epoch": 0.514005983138428,
      "grad_norm": 0.5703125,
      "learning_rate": 1.125248447305569e-05,
      "loss": 0.7586,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3780
    },
    {
      "epoch": 0.5141419635572477,
      "grad_norm": 0.27734375,
      "learning_rate": 1.1247773958947447e-05,
      "loss": 0.5207,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3781
    },
    {
      "epoch": 0.5142779439760674,
      "grad_norm": 0.57421875,
      "learning_rate": 1.1243063163575702e-05,
      "loss": 0.5876,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3782
    },
    {
      "epoch": 0.5144139243948871,
      "grad_norm": 0.482421875,
      "learning_rate": 1.123835208800232e-05,
      "loss": 0.5591,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3783
    },
    {
      "epoch": 0.5145499048137068,
      "grad_norm": 0.279296875,
      "learning_rate": 1.123364073328924e-05,
      "loss": 0.4932,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3784
    },
    {
      "epoch": 0.5146858852325266,
      "grad_norm": 0.306640625,
      "learning_rate": 1.1228929100498453e-05,
      "loss": 0.5202,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3785
    },
    {
      "epoch": 0.5148218656513462,
      "grad_norm": 0.439453125,
      "learning_rate": 1.1224217190692022e-05,
      "loss": 0.5488,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3786
    },
    {
      "epoch": 0.5149578460701659,
      "grad_norm": 0.341796875,
      "learning_rate": 1.121950500493207e-05,
      "loss": 0.6883,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3787
    },
    {
      "epoch": 0.5150938264889856,
      "grad_norm": 0.400390625,
      "learning_rate": 1.1214792544280781e-05,
      "loss": 0.6069,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3788
    },
    {
      "epoch": 0.5152298069078053,
      "grad_norm": 0.392578125,
      "learning_rate": 1.1210079809800397e-05,
      "loss": 0.7044,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3789
    },
    {
      "epoch": 0.515365787326625,
      "grad_norm": 0.314453125,
      "learning_rate": 1.1205366802553231e-05,
      "loss": 0.4863,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3790
    },
    {
      "epoch": 0.5155017677454447,
      "grad_norm": 0.306640625,
      "learning_rate": 1.1200653523601652e-05,
      "loss": 0.4512,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3791
    },
    {
      "epoch": 0.5156377481642643,
      "grad_norm": 0.431640625,
      "learning_rate": 1.1195939974008088e-05,
      "loss": 0.7327,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3792
    },
    {
      "epoch": 0.515773728583084,
      "grad_norm": 0.26953125,
      "learning_rate": 1.1191226154835033e-05,
      "loss": 0.5163,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3793
    },
    {
      "epoch": 0.5159097090019037,
      "grad_norm": 0.4921875,
      "learning_rate": 1.1186512067145039e-05,
      "loss": 0.6289,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3794
    },
    {
      "epoch": 0.5160456894207234,
      "grad_norm": 0.310546875,
      "learning_rate": 1.1181797712000719e-05,
      "loss": 0.6341,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3795
    },
    {
      "epoch": 0.5161816698395431,
      "grad_norm": 0.345703125,
      "learning_rate": 1.1177083090464744e-05,
      "loss": 0.4954,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3796
    },
    {
      "epoch": 0.5163176502583628,
      "grad_norm": 0.353515625,
      "learning_rate": 1.1172368203599853e-05,
      "loss": 0.5055,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3797
    },
    {
      "epoch": 0.5164536306771825,
      "grad_norm": 0.59765625,
      "learning_rate": 1.1167653052468832e-05,
      "loss": 0.5282,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3798
    },
    {
      "epoch": 0.5165896110960022,
      "grad_norm": 0.427734375,
      "learning_rate": 1.1162937638134538e-05,
      "loss": 0.6634,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3799
    },
    {
      "epoch": 0.5167255915148219,
      "grad_norm": 0.3671875,
      "learning_rate": 1.1158221961659883e-05,
      "loss": 0.7186,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3800
    },
    {
      "epoch": 0.5168615719336416,
      "grad_norm": 0.38671875,
      "learning_rate": 1.1153506024107836e-05,
      "loss": 0.6792,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3801
    },
    {
      "epoch": 0.5169975523524613,
      "grad_norm": 0.53515625,
      "learning_rate": 1.1148789826541429e-05,
      "loss": 0.7046,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3802
    },
    {
      "epoch": 0.5171335327712809,
      "grad_norm": 0.283203125,
      "learning_rate": 1.1144073370023751e-05,
      "loss": 0.4119,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3803
    },
    {
      "epoch": 0.5172695131901006,
      "grad_norm": 0.625,
      "learning_rate": 1.1139356655617945e-05,
      "loss": 0.7735,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3804
    },
    {
      "epoch": 0.5174054936089203,
      "grad_norm": 0.423828125,
      "learning_rate": 1.1134639684387222e-05,
      "loss": 0.6986,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3805
    },
    {
      "epoch": 0.51754147402774,
      "grad_norm": 0.80078125,
      "learning_rate": 1.1129922457394842e-05,
      "loss": 0.6807,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3806
    },
    {
      "epoch": 0.5176774544465597,
      "grad_norm": 0.302734375,
      "learning_rate": 1.1125204975704122e-05,
      "loss": 0.6546,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3807
    },
    {
      "epoch": 0.5178134348653793,
      "grad_norm": 0.3671875,
      "learning_rate": 1.1120487240378448e-05,
      "loss": 0.5271,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3808
    },
    {
      "epoch": 0.517949415284199,
      "grad_norm": 0.23828125,
      "learning_rate": 1.1115769252481248e-05,
      "loss": 0.3934,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3809
    },
    {
      "epoch": 0.5180853957030188,
      "grad_norm": 0.55859375,
      "learning_rate": 1.1111051013076017e-05,
      "loss": 0.8613,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3810
    },
    {
      "epoch": 0.5182213761218385,
      "grad_norm": 0.482421875,
      "learning_rate": 1.1106332523226307e-05,
      "loss": 0.7669,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3811
    },
    {
      "epoch": 0.5183573565406582,
      "grad_norm": 0.38671875,
      "learning_rate": 1.1101613783995718e-05,
      "loss": 0.5754,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3812
    },
    {
      "epoch": 0.5184933369594779,
      "grad_norm": 0.62109375,
      "learning_rate": 1.1096894796447912e-05,
      "loss": 0.4733,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3813
    },
    {
      "epoch": 0.5186293173782975,
      "grad_norm": 0.287109375,
      "learning_rate": 1.109217556164661e-05,
      "loss": 0.5373,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3814
    },
    {
      "epoch": 0.5187652977971172,
      "grad_norm": 0.423828125,
      "learning_rate": 1.1087456080655583e-05,
      "loss": 0.6657,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3815
    },
    {
      "epoch": 0.5189012782159369,
      "grad_norm": 0.423828125,
      "learning_rate": 1.108273635453866e-05,
      "loss": 0.7352,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3816
    },
    {
      "epoch": 0.5190372586347566,
      "grad_norm": 0.388671875,
      "learning_rate": 1.1078016384359725e-05,
      "loss": 0.5605,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3817
    },
    {
      "epoch": 0.5191732390535763,
      "grad_norm": 0.69921875,
      "learning_rate": 1.1073296171182716e-05,
      "loss": 0.7045,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3818
    },
    {
      "epoch": 0.5193092194723959,
      "grad_norm": 0.400390625,
      "learning_rate": 1.106857571607163e-05,
      "loss": 0.7743,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3819
    },
    {
      "epoch": 0.5194451998912156,
      "grad_norm": 0.6484375,
      "learning_rate": 1.1063855020090513e-05,
      "loss": 0.8784,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3820
    },
    {
      "epoch": 0.5195811803100353,
      "grad_norm": 0.369140625,
      "learning_rate": 1.1059134084303467e-05,
      "loss": 0.5503,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3821
    },
    {
      "epoch": 0.5197171607288551,
      "grad_norm": 0.310546875,
      "learning_rate": 1.105441290977465e-05,
      "loss": 0.6576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3822
    },
    {
      "epoch": 0.5198531411476748,
      "grad_norm": 0.458984375,
      "learning_rate": 1.104969149756827e-05,
      "loss": 0.5633,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3823
    },
    {
      "epoch": 0.5199891215664945,
      "grad_norm": 0.30078125,
      "learning_rate": 1.1044969848748594e-05,
      "loss": 0.6776,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3824
    },
    {
      "epoch": 0.5201251019853141,
      "grad_norm": 0.306640625,
      "learning_rate": 1.1040247964379938e-05,
      "loss": 0.6047,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3825
    },
    {
      "epoch": 0.5202610824041338,
      "grad_norm": 0.275390625,
      "learning_rate": 1.1035525845526673e-05,
      "loss": 0.4893,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3826
    },
    {
      "epoch": 0.5203970628229535,
      "grad_norm": 0.490234375,
      "learning_rate": 1.103080349325322e-05,
      "loss": 0.6851,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3827
    },
    {
      "epoch": 0.5205330432417732,
      "grad_norm": 0.33984375,
      "learning_rate": 1.102608090862406e-05,
      "loss": 0.7065,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3828
    },
    {
      "epoch": 0.5206690236605929,
      "grad_norm": 0.52734375,
      "learning_rate": 1.1021358092703718e-05,
      "loss": 0.6686,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3829
    },
    {
      "epoch": 0.5208050040794125,
      "grad_norm": 0.44921875,
      "learning_rate": 1.1016635046556773e-05,
      "loss": 0.4887,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3830
    },
    {
      "epoch": 0.5209409844982322,
      "grad_norm": 0.396484375,
      "learning_rate": 1.1011911771247858e-05,
      "loss": 0.6143,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3831
    },
    {
      "epoch": 0.5210769649170519,
      "grad_norm": 0.349609375,
      "learning_rate": 1.1007188267841661e-05,
      "loss": 0.7474,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3832
    },
    {
      "epoch": 0.5212129453358716,
      "grad_norm": 0.33984375,
      "learning_rate": 1.1002464537402913e-05,
      "loss": 0.5887,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3833
    },
    {
      "epoch": 0.5213489257546913,
      "grad_norm": 0.283203125,
      "learning_rate": 1.0997740580996403e-05,
      "loss": 0.5646,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3834
    },
    {
      "epoch": 0.5214849061735111,
      "grad_norm": 0.44921875,
      "learning_rate": 1.0993016399686967e-05,
      "loss": 0.7082,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3835
    },
    {
      "epoch": 0.5216208865923307,
      "grad_norm": 0.419921875,
      "learning_rate": 1.0988291994539493e-05,
      "loss": 0.6965,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3836
    },
    {
      "epoch": 0.5217568670111504,
      "grad_norm": 0.435546875,
      "learning_rate": 1.0983567366618922e-05,
      "loss": 0.6139,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3837
    },
    {
      "epoch": 0.5218928474299701,
      "grad_norm": 0.333984375,
      "learning_rate": 1.097884251699024e-05,
      "loss": 0.5438,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3838
    },
    {
      "epoch": 0.5220288278487898,
      "grad_norm": 0.494140625,
      "learning_rate": 1.0974117446718489e-05,
      "loss": 0.6784,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3839
    },
    {
      "epoch": 0.5221648082676095,
      "grad_norm": 0.38671875,
      "learning_rate": 1.0969392156868754e-05,
      "loss": 0.8081,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3840
    },
    {
      "epoch": 0.5223007886864292,
      "grad_norm": 0.421875,
      "learning_rate": 1.0964666648506175e-05,
      "loss": 0.617,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3841
    },
    {
      "epoch": 0.5224367691052488,
      "grad_norm": 0.294921875,
      "learning_rate": 1.095994092269594e-05,
      "loss": 0.5786,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3842
    },
    {
      "epoch": 0.5225727495240685,
      "grad_norm": 0.56640625,
      "learning_rate": 1.0955214980503284e-05,
      "loss": 0.6204,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3843
    },
    {
      "epoch": 0.5227087299428882,
      "grad_norm": 0.255859375,
      "learning_rate": 1.0950488822993495e-05,
      "loss": 0.5065,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3844
    },
    {
      "epoch": 0.5228447103617079,
      "grad_norm": 0.59375,
      "learning_rate": 1.0945762451231902e-05,
      "loss": 0.6097,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3845
    },
    {
      "epoch": 0.5229806907805276,
      "grad_norm": 0.35546875,
      "learning_rate": 1.0941035866283892e-05,
      "loss": 0.7129,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3846
    },
    {
      "epoch": 0.5231166711993473,
      "grad_norm": 0.2177734375,
      "learning_rate": 1.093630906921489e-05,
      "loss": 0.4232,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3847
    },
    {
      "epoch": 0.523252651618167,
      "grad_norm": 0.38671875,
      "learning_rate": 1.0931582061090377e-05,
      "loss": 0.4854,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3848
    },
    {
      "epoch": 0.5233886320369867,
      "grad_norm": 0.2236328125,
      "learning_rate": 1.0926854842975877e-05,
      "loss": 0.2568,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3849
    },
    {
      "epoch": 0.5235246124558064,
      "grad_norm": 0.390625,
      "learning_rate": 1.0922127415936966e-05,
      "loss": 0.5355,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3850
    },
    {
      "epoch": 0.5236605928746261,
      "grad_norm": 0.259765625,
      "learning_rate": 1.0917399781039259e-05,
      "loss": 0.5353,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3851
    },
    {
      "epoch": 0.5237965732934458,
      "grad_norm": 0.328125,
      "learning_rate": 1.0912671939348428e-05,
      "loss": 0.4437,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3852
    },
    {
      "epoch": 0.5239325537122654,
      "grad_norm": 0.265625,
      "learning_rate": 1.0907943891930181e-05,
      "loss": 0.4996,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3853
    },
    {
      "epoch": 0.5240685341310851,
      "grad_norm": 0.375,
      "learning_rate": 1.0903215639850283e-05,
      "loss": 0.596,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3854
    },
    {
      "epoch": 0.5242045145499048,
      "grad_norm": 0.4609375,
      "learning_rate": 1.0898487184174533e-05,
      "loss": 0.7412,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3855
    },
    {
      "epoch": 0.5243404949687245,
      "grad_norm": 0.294921875,
      "learning_rate": 1.0893758525968789e-05,
      "loss": 0.4181,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3856
    },
    {
      "epoch": 0.5244764753875442,
      "grad_norm": 0.314453125,
      "learning_rate": 1.0889029666298947e-05,
      "loss": 0.603,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3857
    },
    {
      "epoch": 0.5246124558063638,
      "grad_norm": 0.5234375,
      "learning_rate": 1.0884300606230949e-05,
      "loss": 0.5799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3858
    },
    {
      "epoch": 0.5247484362251835,
      "grad_norm": 0.482421875,
      "learning_rate": 1.087957134683078e-05,
      "loss": 0.5814,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3859
    },
    {
      "epoch": 0.5248844166440033,
      "grad_norm": 0.482421875,
      "learning_rate": 1.0874841889164478e-05,
      "loss": 0.7645,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3860
    },
    {
      "epoch": 0.525020397062823,
      "grad_norm": 0.33203125,
      "learning_rate": 1.0870112234298118e-05,
      "loss": 0.5365,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3861
    },
    {
      "epoch": 0.5251563774816427,
      "grad_norm": 0.26953125,
      "learning_rate": 1.0865382383297823e-05,
      "loss": 0.3485,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3862
    },
    {
      "epoch": 0.5252923579004624,
      "grad_norm": 0.36328125,
      "learning_rate": 1.0860652337229755e-05,
      "loss": 0.757,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3863
    },
    {
      "epoch": 0.525428338319282,
      "grad_norm": 0.3046875,
      "learning_rate": 1.0855922097160131e-05,
      "loss": 0.5223,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3864
    },
    {
      "epoch": 0.5255643187381017,
      "grad_norm": 0.6875,
      "learning_rate": 1.0851191664155202e-05,
      "loss": 0.674,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3865
    },
    {
      "epoch": 0.5257002991569214,
      "grad_norm": 0.40625,
      "learning_rate": 1.0846461039281265e-05,
      "loss": 0.6564,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3866
    },
    {
      "epoch": 0.5258362795757411,
      "grad_norm": 0.412109375,
      "learning_rate": 1.0841730223604659e-05,
      "loss": 0.8179,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3867
    },
    {
      "epoch": 0.5259722599945608,
      "grad_norm": 0.546875,
      "learning_rate": 1.0836999218191771e-05,
      "loss": 0.9536,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3868
    },
    {
      "epoch": 0.5261082404133804,
      "grad_norm": 0.33984375,
      "learning_rate": 1.0832268024109025e-05,
      "loss": 0.6976,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3869
    },
    {
      "epoch": 0.5262442208322001,
      "grad_norm": 0.333984375,
      "learning_rate": 1.0827536642422894e-05,
      "loss": 0.6587,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3870
    },
    {
      "epoch": 0.5263802012510198,
      "grad_norm": 0.59375,
      "learning_rate": 1.0822805074199882e-05,
      "loss": 0.6742,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3871
    },
    {
      "epoch": 0.5265161816698395,
      "grad_norm": 0.39453125,
      "learning_rate": 1.0818073320506547e-05,
      "loss": 0.5993,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3872
    },
    {
      "epoch": 0.5266521620886593,
      "grad_norm": 0.337890625,
      "learning_rate": 1.0813341382409488e-05,
      "loss": 0.6086,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3873
    },
    {
      "epoch": 0.526788142507479,
      "grad_norm": 0.82421875,
      "learning_rate": 1.0808609260975333e-05,
      "loss": 0.625,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3874
    },
    {
      "epoch": 0.5269241229262986,
      "grad_norm": 0.3046875,
      "learning_rate": 1.0803876957270766e-05,
      "loss": 0.4312,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3875
    },
    {
      "epoch": 0.5270601033451183,
      "grad_norm": 0.61328125,
      "learning_rate": 1.0799144472362506e-05,
      "loss": 0.7194,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3876
    },
    {
      "epoch": 0.527196083763938,
      "grad_norm": 0.40234375,
      "learning_rate": 1.0794411807317312e-05,
      "loss": 0.7563,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3877
    },
    {
      "epoch": 0.5273320641827577,
      "grad_norm": 0.625,
      "learning_rate": 1.0789678963201982e-05,
      "loss": 0.624,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3878
    },
    {
      "epoch": 0.5274680446015774,
      "grad_norm": 0.6484375,
      "learning_rate": 1.0784945941083358e-05,
      "loss": 0.6683,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3879
    },
    {
      "epoch": 0.527604025020397,
      "grad_norm": 0.490234375,
      "learning_rate": 1.0780212742028324e-05,
      "loss": 0.8656,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3880
    },
    {
      "epoch": 0.5277400054392167,
      "grad_norm": 0.353515625,
      "learning_rate": 1.0775479367103798e-05,
      "loss": 0.5485,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3881
    },
    {
      "epoch": 0.5278759858580364,
      "grad_norm": 0.40625,
      "learning_rate": 1.0770745817376741e-05,
      "loss": 0.6626,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3882
    },
    {
      "epoch": 0.5280119662768561,
      "grad_norm": 2.828125,
      "learning_rate": 1.0766012093914156e-05,
      "loss": 0.6289,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3883
    },
    {
      "epoch": 0.5281479466956758,
      "grad_norm": 0.28515625,
      "learning_rate": 1.0761278197783079e-05,
      "loss": 0.5542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3884
    },
    {
      "epoch": 0.5282839271144956,
      "grad_norm": 0.53515625,
      "learning_rate": 1.0756544130050586e-05,
      "loss": 0.9645,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3885
    },
    {
      "epoch": 0.5284199075333152,
      "grad_norm": 0.41796875,
      "learning_rate": 1.0751809891783799e-05,
      "loss": 0.6099,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3886
    },
    {
      "epoch": 0.5285558879521349,
      "grad_norm": 0.396484375,
      "learning_rate": 1.0747075484049872e-05,
      "loss": 0.7419,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3887
    },
    {
      "epoch": 0.5286918683709546,
      "grad_norm": 0.447265625,
      "learning_rate": 1.0742340907915993e-05,
      "loss": 0.7349,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3888
    },
    {
      "epoch": 0.5288278487897743,
      "grad_norm": 0.205078125,
      "learning_rate": 1.07376061644494e-05,
      "loss": 0.3188,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3889
    },
    {
      "epoch": 0.528963829208594,
      "grad_norm": 0.365234375,
      "learning_rate": 1.073287125471736e-05,
      "loss": 0.7174,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3890
    },
    {
      "epoch": 0.5290998096274137,
      "grad_norm": 0.2578125,
      "learning_rate": 1.0728136179787178e-05,
      "loss": 0.4561,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3891
    },
    {
      "epoch": 0.5292357900462333,
      "grad_norm": 0.45703125,
      "learning_rate": 1.07234009407262e-05,
      "loss": 0.7082,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3892
    },
    {
      "epoch": 0.529371770465053,
      "grad_norm": 0.5234375,
      "learning_rate": 1.0718665538601804e-05,
      "loss": 0.5298,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3893
    },
    {
      "epoch": 0.5295077508838727,
      "grad_norm": 0.26953125,
      "learning_rate": 1.071392997448141e-05,
      "loss": 0.465,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3894
    },
    {
      "epoch": 0.5296437313026924,
      "grad_norm": 0.3359375,
      "learning_rate": 1.0709194249432471e-05,
      "loss": 0.6157,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3895
    },
    {
      "epoch": 0.5297797117215121,
      "grad_norm": 0.435546875,
      "learning_rate": 1.070445836452248e-05,
      "loss": 0.7625,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3896
    },
    {
      "epoch": 0.5299156921403317,
      "grad_norm": 0.330078125,
      "learning_rate": 1.0699722320818957e-05,
      "loss": 0.6764,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3897
    },
    {
      "epoch": 0.5300516725591515,
      "grad_norm": 0.33984375,
      "learning_rate": 1.069498611938947e-05,
      "loss": 0.652,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3898
    },
    {
      "epoch": 0.5301876529779712,
      "grad_norm": 0.62109375,
      "learning_rate": 1.0690249761301614e-05,
      "loss": 0.6825,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3899
    },
    {
      "epoch": 0.5303236333967909,
      "grad_norm": 0.28515625,
      "learning_rate": 1.0685513247623023e-05,
      "loss": 0.585,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3900
    },
    {
      "epoch": 0.5304596138156106,
      "grad_norm": 0.353515625,
      "learning_rate": 1.0680776579421363e-05,
      "loss": 0.6341,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3901
    },
    {
      "epoch": 0.5305955942344303,
      "grad_norm": 0.37890625,
      "learning_rate": 1.0676039757764342e-05,
      "loss": 0.7407,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3902
    },
    {
      "epoch": 0.5307315746532499,
      "grad_norm": 0.80078125,
      "learning_rate": 1.0671302783719692e-05,
      "loss": 0.8491,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3903
    },
    {
      "epoch": 0.5308675550720696,
      "grad_norm": 0.26953125,
      "learning_rate": 1.0666565658355187e-05,
      "loss": 0.5996,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3904
    },
    {
      "epoch": 0.5310035354908893,
      "grad_norm": 0.41796875,
      "learning_rate": 1.0661828382738637e-05,
      "loss": 0.5622,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3905
    },
    {
      "epoch": 0.531139515909709,
      "grad_norm": 0.44140625,
      "learning_rate": 1.0657090957937874e-05,
      "loss": 0.6578,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3906
    },
    {
      "epoch": 0.5312754963285287,
      "grad_norm": 0.443359375,
      "learning_rate": 1.0652353385020777e-05,
      "loss": 0.6932,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3907
    },
    {
      "epoch": 0.5314114767473483,
      "grad_norm": 0.49609375,
      "learning_rate": 1.0647615665055251e-05,
      "loss": 0.5837,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3908
    },
    {
      "epoch": 0.531547457166168,
      "grad_norm": 0.31640625,
      "learning_rate": 1.0642877799109234e-05,
      "loss": 0.5329,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3909
    },
    {
      "epoch": 0.5316834375849878,
      "grad_norm": 0.330078125,
      "learning_rate": 1.0638139788250703e-05,
      "loss": 0.7259,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3910
    },
    {
      "epoch": 0.5318194180038075,
      "grad_norm": 0.2578125,
      "learning_rate": 1.0633401633547663e-05,
      "loss": 0.4451,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3911
    },
    {
      "epoch": 0.5319553984226272,
      "grad_norm": 0.341796875,
      "learning_rate": 1.0628663336068148e-05,
      "loss": 0.5413,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3912
    },
    {
      "epoch": 0.5320913788414469,
      "grad_norm": 0.439453125,
      "learning_rate": 1.0623924896880234e-05,
      "loss": 0.7988,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3913
    },
    {
      "epoch": 0.5322273592602665,
      "grad_norm": 0.458984375,
      "learning_rate": 1.061918631705202e-05,
      "loss": 0.8092,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3914
    },
    {
      "epoch": 0.5323633396790862,
      "grad_norm": 0.609375,
      "learning_rate": 1.0614447597651638e-05,
      "loss": 0.6973,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3915
    },
    {
      "epoch": 0.5324993200979059,
      "grad_norm": 0.27734375,
      "learning_rate": 1.0609708739747256e-05,
      "loss": 0.4943,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3916
    },
    {
      "epoch": 0.5326353005167256,
      "grad_norm": 0.326171875,
      "learning_rate": 1.060496974440707e-05,
      "loss": 0.6167,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3917
    },
    {
      "epoch": 0.5327712809355453,
      "grad_norm": 0.7578125,
      "learning_rate": 1.0600230612699306e-05,
      "loss": 0.7939,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3918
    },
    {
      "epoch": 0.532907261354365,
      "grad_norm": 0.40625,
      "learning_rate": 1.0595491345692226e-05,
      "loss": 0.9072,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3919
    },
    {
      "epoch": 0.5330432417731846,
      "grad_norm": 0.68359375,
      "learning_rate": 1.0590751944454114e-05,
      "loss": 0.8514,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3920
    },
    {
      "epoch": 0.5331792221920043,
      "grad_norm": 0.37890625,
      "learning_rate": 1.0586012410053293e-05,
      "loss": 0.6803,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3921
    },
    {
      "epoch": 0.533315202610824,
      "grad_norm": 0.326171875,
      "learning_rate": 1.0581272743558106e-05,
      "loss": 0.7028,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3922
    },
    {
      "epoch": 0.5334511830296438,
      "grad_norm": 0.5546875,
      "learning_rate": 1.0576532946036938e-05,
      "loss": 0.6618,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3923
    },
    {
      "epoch": 0.5335871634484635,
      "grad_norm": 0.54296875,
      "learning_rate": 1.0571793018558199e-05,
      "loss": 0.6351,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3924
    },
    {
      "epoch": 0.5337231438672831,
      "grad_norm": 0.310546875,
      "learning_rate": 1.0567052962190318e-05,
      "loss": 0.6483,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3925
    },
    {
      "epoch": 0.5338591242861028,
      "grad_norm": 0.318359375,
      "learning_rate": 1.056231277800177e-05,
      "loss": 0.3605,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3926
    },
    {
      "epoch": 0.5339951047049225,
      "grad_norm": 0.330078125,
      "learning_rate": 1.0557572467061044e-05,
      "loss": 0.6125,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3927
    },
    {
      "epoch": 0.5341310851237422,
      "grad_norm": 0.421875,
      "learning_rate": 1.0552832030436668e-05,
      "loss": 0.6738,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3928
    },
    {
      "epoch": 0.5342670655425619,
      "grad_norm": 0.341796875,
      "learning_rate": 1.0548091469197194e-05,
      "loss": 0.6291,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3929
    },
    {
      "epoch": 0.5344030459613816,
      "grad_norm": 0.65234375,
      "learning_rate": 1.0543350784411201e-05,
      "loss": 0.6253,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3930
    },
    {
      "epoch": 0.5345390263802012,
      "grad_norm": 0.625,
      "learning_rate": 1.05386099771473e-05,
      "loss": 0.7451,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3931
    },
    {
      "epoch": 0.5346750067990209,
      "grad_norm": 0.5703125,
      "learning_rate": 1.0533869048474123e-05,
      "loss": 0.5508,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3932
    },
    {
      "epoch": 0.5348109872178406,
      "grad_norm": 0.41796875,
      "learning_rate": 1.0529127999460335e-05,
      "loss": 0.7516,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3933
    },
    {
      "epoch": 0.5349469676366603,
      "grad_norm": 0.31640625,
      "learning_rate": 1.0524386831174629e-05,
      "loss": 0.6452,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3934
    },
    {
      "epoch": 0.53508294805548,
      "grad_norm": 0.3359375,
      "learning_rate": 1.0519645544685719e-05,
      "loss": 0.5462,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3935
    },
    {
      "epoch": 0.5352189284742997,
      "grad_norm": 0.3046875,
      "learning_rate": 1.051490414106235e-05,
      "loss": 0.5653,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3936
    },
    {
      "epoch": 0.5353549088931194,
      "grad_norm": 0.283203125,
      "learning_rate": 1.0510162621373295e-05,
      "loss": 0.6198,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3937
    },
    {
      "epoch": 0.5354908893119391,
      "grad_norm": 0.37109375,
      "learning_rate": 1.0505420986687344e-05,
      "loss": 0.6242,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3938
    },
    {
      "epoch": 0.5356268697307588,
      "grad_norm": 0.4921875,
      "learning_rate": 1.0500679238073326e-05,
      "loss": 0.7417,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3939
    },
    {
      "epoch": 0.5357628501495785,
      "grad_norm": 0.421875,
      "learning_rate": 1.0495937376600081e-05,
      "loss": 0.7161,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3940
    },
    {
      "epoch": 0.5358988305683982,
      "grad_norm": 0.4921875,
      "learning_rate": 1.0491195403336491e-05,
      "loss": 0.7384,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3941
    },
    {
      "epoch": 0.5360348109872178,
      "grad_norm": 0.4296875,
      "learning_rate": 1.0486453319351454e-05,
      "loss": 0.8078,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3942
    },
    {
      "epoch": 0.5361707914060375,
      "grad_norm": 0.625,
      "learning_rate": 1.0481711125713889e-05,
      "loss": 0.7179,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3943
    },
    {
      "epoch": 0.5363067718248572,
      "grad_norm": 0.58984375,
      "learning_rate": 1.0476968823492749e-05,
      "loss": 0.7062,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3944
    },
    {
      "epoch": 0.5364427522436769,
      "grad_norm": 0.9765625,
      "learning_rate": 1.0472226413757003e-05,
      "loss": 0.743,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3945
    },
    {
      "epoch": 0.5365787326624966,
      "grad_norm": 0.59375,
      "learning_rate": 1.046748389757565e-05,
      "loss": 0.7988,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3946
    },
    {
      "epoch": 0.5367147130813162,
      "grad_norm": 0.3359375,
      "learning_rate": 1.0462741276017711e-05,
      "loss": 0.7279,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3947
    },
    {
      "epoch": 0.536850693500136,
      "grad_norm": 0.796875,
      "learning_rate": 1.0457998550152231e-05,
      "loss": 0.6704,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3948
    },
    {
      "epoch": 0.5369866739189557,
      "grad_norm": 0.9296875,
      "learning_rate": 1.045325572104828e-05,
      "loss": 0.8242,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3949
    },
    {
      "epoch": 0.5371226543377754,
      "grad_norm": 0.54296875,
      "learning_rate": 1.0448512789774944e-05,
      "loss": 0.4077,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3950
    },
    {
      "epoch": 0.5372586347565951,
      "grad_norm": 0.376953125,
      "learning_rate": 1.0443769757401346e-05,
      "loss": 0.6667,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3951
    },
    {
      "epoch": 0.5373946151754148,
      "grad_norm": 0.494140625,
      "learning_rate": 1.0439026624996615e-05,
      "loss": 0.7303,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3952
    },
    {
      "epoch": 0.5375305955942344,
      "grad_norm": 1.796875,
      "learning_rate": 1.0434283393629918e-05,
      "loss": 1.0303,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3953
    },
    {
      "epoch": 0.5376665760130541,
      "grad_norm": 0.412109375,
      "learning_rate": 1.0429540064370433e-05,
      "loss": 0.5605,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3954
    },
    {
      "epoch": 0.5378025564318738,
      "grad_norm": 0.365234375,
      "learning_rate": 1.0424796638287367e-05,
      "loss": 0.561,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3955
    },
    {
      "epoch": 0.5379385368506935,
      "grad_norm": 0.76171875,
      "learning_rate": 1.0420053116449944e-05,
      "loss": 0.7869,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3956
    },
    {
      "epoch": 0.5380745172695132,
      "grad_norm": 0.40234375,
      "learning_rate": 1.0415309499927415e-05,
      "loss": 0.6405,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3957
    },
    {
      "epoch": 0.5382104976883328,
      "grad_norm": 0.337890625,
      "learning_rate": 1.0410565789789044e-05,
      "loss": 0.6222,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3958
    },
    {
      "epoch": 0.5383464781071525,
      "grad_norm": 0.60546875,
      "learning_rate": 1.0405821987104128e-05,
      "loss": 0.5444,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3959
    },
    {
      "epoch": 0.5384824585259722,
      "grad_norm": 0.51171875,
      "learning_rate": 1.0401078092941972e-05,
      "loss": 0.32,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3960
    },
    {
      "epoch": 0.538618438944792,
      "grad_norm": 0.66015625,
      "learning_rate": 1.039633410837191e-05,
      "loss": 0.7965,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3961
    },
    {
      "epoch": 0.5387544193636117,
      "grad_norm": 0.458984375,
      "learning_rate": 1.0391590034463295e-05,
      "loss": 0.7186,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3962
    },
    {
      "epoch": 0.5388903997824314,
      "grad_norm": 0.240234375,
      "learning_rate": 1.03868458722855e-05,
      "loss": 0.4222,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3963
    },
    {
      "epoch": 0.539026380201251,
      "grad_norm": 0.5234375,
      "learning_rate": 1.0382101622907914e-05,
      "loss": 0.7902,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3964
    },
    {
      "epoch": 0.5391623606200707,
      "grad_norm": 0.38671875,
      "learning_rate": 1.0377357287399952e-05,
      "loss": 0.7197,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3965
    },
    {
      "epoch": 0.5392983410388904,
      "grad_norm": 0.28515625,
      "learning_rate": 1.0372612866831042e-05,
      "loss": 0.6113,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3966
    },
    {
      "epoch": 0.5394343214577101,
      "grad_norm": 0.5546875,
      "learning_rate": 1.0367868362270639e-05,
      "loss": 0.8021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3967
    },
    {
      "epoch": 0.5395703018765298,
      "grad_norm": 0.30859375,
      "learning_rate": 1.0363123774788206e-05,
      "loss": 0.6143,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3968
    },
    {
      "epoch": 0.5397062822953494,
      "grad_norm": 0.4609375,
      "learning_rate": 1.0358379105453239e-05,
      "loss": 0.6806,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3969
    },
    {
      "epoch": 0.5398422627141691,
      "grad_norm": 0.53125,
      "learning_rate": 1.0353634355335234e-05,
      "loss": 0.6118,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3970
    },
    {
      "epoch": 0.5399782431329888,
      "grad_norm": 0.400390625,
      "learning_rate": 1.0348889525503725e-05,
      "loss": 0.5378,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3971
    },
    {
      "epoch": 0.5401142235518085,
      "grad_norm": 0.462890625,
      "learning_rate": 1.034414461702825e-05,
      "loss": 0.6626,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3972
    },
    {
      "epoch": 0.5402502039706283,
      "grad_norm": 0.3984375,
      "learning_rate": 1.0339399630978373e-05,
      "loss": 0.7798,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3973
    },
    {
      "epoch": 0.540386184389448,
      "grad_norm": 0.41796875,
      "learning_rate": 1.033465456842367e-05,
      "loss": 0.6217,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3974
    },
    {
      "epoch": 0.5405221648082676,
      "grad_norm": 0.357421875,
      "learning_rate": 1.0329909430433733e-05,
      "loss": 0.4829,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3975
    },
    {
      "epoch": 0.5406581452270873,
      "grad_norm": 0.37890625,
      "learning_rate": 1.032516421807818e-05,
      "loss": 0.6428,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3976
    },
    {
      "epoch": 0.540794125645907,
      "grad_norm": 0.1884765625,
      "learning_rate": 1.0320418932426634e-05,
      "loss": 0.3923,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3977
    },
    {
      "epoch": 0.5409301060647267,
      "grad_norm": 0.421875,
      "learning_rate": 1.0315673574548746e-05,
      "loss": 0.5282,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3978
    },
    {
      "epoch": 0.5410660864835464,
      "grad_norm": 0.318359375,
      "learning_rate": 1.0310928145514173e-05,
      "loss": 0.6077,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3979
    },
    {
      "epoch": 0.541202066902366,
      "grad_norm": 0.490234375,
      "learning_rate": 1.0306182646392594e-05,
      "loss": 0.6942,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3980
    },
    {
      "epoch": 0.5413380473211857,
      "grad_norm": 0.53515625,
      "learning_rate": 1.0301437078253707e-05,
      "loss": 0.8532,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3981
    },
    {
      "epoch": 0.5414740277400054,
      "grad_norm": 0.271484375,
      "learning_rate": 1.0296691442167212e-05,
      "loss": 0.5358,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3982
    },
    {
      "epoch": 0.5416100081588251,
      "grad_norm": 0.58203125,
      "learning_rate": 1.0291945739202843e-05,
      "loss": 0.5859,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3983
    },
    {
      "epoch": 0.5417459885776448,
      "grad_norm": 0.53125,
      "learning_rate": 1.0287199970430331e-05,
      "loss": 0.8643,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3984
    },
    {
      "epoch": 0.5418819689964645,
      "grad_norm": 0.349609375,
      "learning_rate": 1.0282454136919438e-05,
      "loss": 0.6543,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3985
    },
    {
      "epoch": 0.5420179494152843,
      "grad_norm": 0.3515625,
      "learning_rate": 1.0277708239739925e-05,
      "loss": 0.5866,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3986
    },
    {
      "epoch": 0.5421539298341039,
      "grad_norm": 3.015625,
      "learning_rate": 1.0272962279961583e-05,
      "loss": 0.8009,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3987
    },
    {
      "epoch": 0.5422899102529236,
      "grad_norm": 0.294921875,
      "learning_rate": 1.0268216258654204e-05,
      "loss": 0.4469,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3988
    },
    {
      "epoch": 0.5424258906717433,
      "grad_norm": 0.50390625,
      "learning_rate": 1.0263470176887602e-05,
      "loss": 0.71,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3989
    },
    {
      "epoch": 0.542561871090563,
      "grad_norm": 0.36328125,
      "learning_rate": 1.0258724035731602e-05,
      "loss": 0.5049,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3990
    },
    {
      "epoch": 0.5426978515093827,
      "grad_norm": 0.248046875,
      "learning_rate": 1.0253977836256041e-05,
      "loss": 0.5116,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3991
    },
    {
      "epoch": 0.5428338319282023,
      "grad_norm": 0.45703125,
      "learning_rate": 1.0249231579530769e-05,
      "loss": 0.7915,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3992
    },
    {
      "epoch": 0.542969812347022,
      "grad_norm": 0.6875,
      "learning_rate": 1.0244485266625652e-05,
      "loss": 0.5161,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3993
    },
    {
      "epoch": 0.5431057927658417,
      "grad_norm": 0.33984375,
      "learning_rate": 1.0239738898610569e-05,
      "loss": 0.6112,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3994
    },
    {
      "epoch": 0.5432417731846614,
      "grad_norm": 0.318359375,
      "learning_rate": 1.0234992476555406e-05,
      "loss": 0.578,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3995
    },
    {
      "epoch": 0.5433777536034811,
      "grad_norm": 0.287109375,
      "learning_rate": 1.0230246001530065e-05,
      "loss": 0.5176,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3996
    },
    {
      "epoch": 0.5435137340223007,
      "grad_norm": 0.431640625,
      "learning_rate": 1.0225499474604459e-05,
      "loss": 0.6855,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3997
    },
    {
      "epoch": 0.5436497144411204,
      "grad_norm": 0.2158203125,
      "learning_rate": 1.0220752896848516e-05,
      "loss": 0.4141,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3998
    },
    {
      "epoch": 0.5437856948599402,
      "grad_norm": 0.326171875,
      "learning_rate": 1.021600626933217e-05,
      "loss": 0.6696,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 3999
    },
    {
      "epoch": 0.5439216752787599,
      "grad_norm": 0.40625,
      "learning_rate": 1.021125959312537e-05,
      "loss": 0.7953,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4000
    },
    {
      "epoch": 0.5440576556975796,
      "grad_norm": 0.232421875,
      "learning_rate": 1.0206512869298075e-05,
      "loss": 0.3942,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4001
    },
    {
      "epoch": 0.5441936361163993,
      "grad_norm": 0.357421875,
      "learning_rate": 1.0201766098920256e-05,
      "loss": 0.5529,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4002
    },
    {
      "epoch": 0.5443296165352189,
      "grad_norm": 0.5859375,
      "learning_rate": 1.0197019283061888e-05,
      "loss": 0.6716,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4003
    },
    {
      "epoch": 0.5444655969540386,
      "grad_norm": 0.421875,
      "learning_rate": 1.0192272422792968e-05,
      "loss": 0.7103,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4004
    },
    {
      "epoch": 0.5446015773728583,
      "grad_norm": 0.3203125,
      "learning_rate": 1.0187525519183494e-05,
      "loss": 0.689,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4005
    },
    {
      "epoch": 0.544737557791678,
      "grad_norm": 0.39453125,
      "learning_rate": 1.0182778573303473e-05,
      "loss": 0.6117,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4006
    },
    {
      "epoch": 0.5448735382104977,
      "grad_norm": 0.326171875,
      "learning_rate": 1.0178031586222927e-05,
      "loss": 0.6258,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4007
    },
    {
      "epoch": 0.5450095186293173,
      "grad_norm": 0.54296875,
      "learning_rate": 1.0173284559011886e-05,
      "loss": 0.812,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4008
    },
    {
      "epoch": 0.545145499048137,
      "grad_norm": 0.72265625,
      "learning_rate": 1.0168537492740387e-05,
      "loss": 0.8203,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4009
    },
    {
      "epoch": 0.5452814794669567,
      "grad_norm": 0.37890625,
      "learning_rate": 1.0163790388478476e-05,
      "loss": 0.6855,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4010
    },
    {
      "epoch": 0.5454174598857765,
      "grad_norm": 0.6328125,
      "learning_rate": 1.015904324729621e-05,
      "loss": 0.5991,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4011
    },
    {
      "epoch": 0.5455534403045962,
      "grad_norm": 0.423828125,
      "learning_rate": 1.0154296070263649e-05,
      "loss": 0.7508,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4012
    },
    {
      "epoch": 0.5456894207234159,
      "grad_norm": 0.228515625,
      "learning_rate": 1.0149548858450871e-05,
      "loss": 0.5124,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4013
    },
    {
      "epoch": 0.5458254011422355,
      "grad_norm": 0.34375,
      "learning_rate": 1.0144801612927948e-05,
      "loss": 0.6241,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4014
    },
    {
      "epoch": 0.5459613815610552,
      "grad_norm": 0.2255859375,
      "learning_rate": 1.0140054334764974e-05,
      "loss": 0.4373,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4015
    },
    {
      "epoch": 0.5460973619798749,
      "grad_norm": 0.462890625,
      "learning_rate": 1.013530702503204e-05,
      "loss": 0.6986,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4016
    },
    {
      "epoch": 0.5462333423986946,
      "grad_norm": 0.390625,
      "learning_rate": 1.0130559684799247e-05,
      "loss": 0.7274,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4017
    },
    {
      "epoch": 0.5463693228175143,
      "grad_norm": 0.75,
      "learning_rate": 1.0125812315136706e-05,
      "loss": 0.5562,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4018
    },
    {
      "epoch": 0.546505303236334,
      "grad_norm": 0.6875,
      "learning_rate": 1.0121064917114529e-05,
      "loss": 0.6951,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4019
    },
    {
      "epoch": 0.5466412836551536,
      "grad_norm": 0.6015625,
      "learning_rate": 1.011631749180284e-05,
      "loss": 0.638,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4020
    },
    {
      "epoch": 0.5467772640739733,
      "grad_norm": 0.25390625,
      "learning_rate": 1.0111570040271768e-05,
      "loss": 0.3813,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4021
    },
    {
      "epoch": 0.546913244492793,
      "grad_norm": 0.435546875,
      "learning_rate": 1.0106822563591445e-05,
      "loss": 0.6209,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4022
    },
    {
      "epoch": 0.5470492249116127,
      "grad_norm": 0.263671875,
      "learning_rate": 1.0102075062832009e-05,
      "loss": 0.4585,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4023
    },
    {
      "epoch": 0.5471852053304325,
      "grad_norm": 0.396484375,
      "learning_rate": 1.0097327539063605e-05,
      "loss": 0.744,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4024
    },
    {
      "epoch": 0.5473211857492521,
      "grad_norm": 0.42578125,
      "learning_rate": 1.0092579993356386e-05,
      "loss": 0.6647,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4025
    },
    {
      "epoch": 0.5474571661680718,
      "grad_norm": 0.75390625,
      "learning_rate": 1.0087832426780504e-05,
      "loss": 0.762,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4026
    },
    {
      "epoch": 0.5475931465868915,
      "grad_norm": 0.5,
      "learning_rate": 1.0083084840406119e-05,
      "loss": 0.7295,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4027
    },
    {
      "epoch": 0.5477291270057112,
      "grad_norm": 0.60546875,
      "learning_rate": 1.0078337235303396e-05,
      "loss": 0.7795,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4028
    },
    {
      "epoch": 0.5478651074245309,
      "grad_norm": 0.373046875,
      "learning_rate": 1.00735896125425e-05,
      "loss": 0.5564,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4029
    },
    {
      "epoch": 0.5480010878433506,
      "grad_norm": 0.54296875,
      "learning_rate": 1.0068841973193604e-05,
      "loss": 0.7176,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4030
    },
    {
      "epoch": 0.5481370682621702,
      "grad_norm": 0.4921875,
      "learning_rate": 1.0064094318326888e-05,
      "loss": 0.5215,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4031
    },
    {
      "epoch": 0.5482730486809899,
      "grad_norm": 0.3125,
      "learning_rate": 1.0059346649012529e-05,
      "loss": 0.5073,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4032
    },
    {
      "epoch": 0.5484090290998096,
      "grad_norm": 0.4296875,
      "learning_rate": 1.0054598966320712e-05,
      "loss": 0.798,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4033
    },
    {
      "epoch": 0.5485450095186293,
      "grad_norm": 0.244140625,
      "learning_rate": 1.0049851271321618e-05,
      "loss": 0.498,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4034
    },
    {
      "epoch": 0.548680989937449,
      "grad_norm": 0.30078125,
      "learning_rate": 1.0045103565085438e-05,
      "loss": 0.5417,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4035
    },
    {
      "epoch": 0.5488169703562688,
      "grad_norm": 0.31640625,
      "learning_rate": 1.0040355848682364e-05,
      "loss": 0.5309,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4036
    },
    {
      "epoch": 0.5489529507750884,
      "grad_norm": 0.5703125,
      "learning_rate": 1.0035608123182588e-05,
      "loss": 0.8174,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4037
    },
    {
      "epoch": 0.5490889311939081,
      "grad_norm": 0.32421875,
      "learning_rate": 1.0030860389656306e-05,
      "loss": 0.4373,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4038
    },
    {
      "epoch": 0.5492249116127278,
      "grad_norm": 0.287109375,
      "learning_rate": 1.0026112649173713e-05,
      "loss": 0.5634,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4039
    },
    {
      "epoch": 0.5493608920315475,
      "grad_norm": 0.5,
      "learning_rate": 1.0021364902805014e-05,
      "loss": 0.8991,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4040
    },
    {
      "epoch": 0.5494968724503672,
      "grad_norm": 0.62109375,
      "learning_rate": 1.00166171516204e-05,
      "loss": 0.6829,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4041
    },
    {
      "epoch": 0.5496328528691868,
      "grad_norm": 0.388671875,
      "learning_rate": 1.001186939669008e-05,
      "loss": 0.7223,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4042
    },
    {
      "epoch": 0.5497688332880065,
      "grad_norm": 0.3828125,
      "learning_rate": 1.000712163908425e-05,
      "loss": 0.5729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4043
    },
    {
      "epoch": 0.5499048137068262,
      "grad_norm": 0.384765625,
      "learning_rate": 1.0002373879873118e-05,
      "loss": 0.5999,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4044
    },
    {
      "epoch": 0.5500407941256459,
      "grad_norm": 0.359375,
      "learning_rate": 9.997626120126885e-06,
      "loss": 0.5566,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4045
    },
    {
      "epoch": 0.5501767745444656,
      "grad_norm": 0.416015625,
      "learning_rate": 9.992878360915752e-06,
      "loss": 0.736,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4046
    },
    {
      "epoch": 0.5503127549632852,
      "grad_norm": 0.5,
      "learning_rate": 9.988130603309924e-06,
      "loss": 0.6559,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4047
    },
    {
      "epoch": 0.5504487353821049,
      "grad_norm": 0.70703125,
      "learning_rate": 9.983382848379602e-06,
      "loss": 0.7225,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4048
    },
    {
      "epoch": 0.5505847158009247,
      "grad_norm": 0.3125,
      "learning_rate": 9.978635097194991e-06,
      "loss": 0.4549,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4049
    },
    {
      "epoch": 0.5507206962197444,
      "grad_norm": 0.478515625,
      "learning_rate": 9.973887350826289e-06,
      "loss": 0.7248,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4050
    },
    {
      "epoch": 0.5508566766385641,
      "grad_norm": 0.3671875,
      "learning_rate": 9.969139610343696e-06,
      "loss": 0.7899,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4051
    },
    {
      "epoch": 0.5509926570573838,
      "grad_norm": 0.4140625,
      "learning_rate": 9.964391876817413e-06,
      "loss": 0.6403,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4052
    },
    {
      "epoch": 0.5511286374762034,
      "grad_norm": 0.466796875,
      "learning_rate": 9.959644151317639e-06,
      "loss": 0.7433,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4053
    },
    {
      "epoch": 0.5512646178950231,
      "grad_norm": 0.38671875,
      "learning_rate": 9.954896434914563e-06,
      "loss": 0.6927,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4054
    },
    {
      "epoch": 0.5514005983138428,
      "grad_norm": 0.33203125,
      "learning_rate": 9.950148728678385e-06,
      "loss": 0.6221,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4055
    },
    {
      "epoch": 0.5515365787326625,
      "grad_norm": 0.427734375,
      "learning_rate": 9.945401033679291e-06,
      "loss": 0.6364,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4056
    },
    {
      "epoch": 0.5516725591514822,
      "grad_norm": 0.28515625,
      "learning_rate": 9.940653350987473e-06,
      "loss": 0.4832,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4057
    },
    {
      "epoch": 0.5518085395703018,
      "grad_norm": 0.6875,
      "learning_rate": 9.935905681673115e-06,
      "loss": 0.763,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4058
    },
    {
      "epoch": 0.5519445199891215,
      "grad_norm": 0.609375,
      "learning_rate": 9.931158026806398e-06,
      "loss": 0.5938,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4059
    },
    {
      "epoch": 0.5520805004079412,
      "grad_norm": 0.375,
      "learning_rate": 9.926410387457506e-06,
      "loss": 0.479,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4060
    },
    {
      "epoch": 0.5522164808267609,
      "grad_norm": 0.3515625,
      "learning_rate": 9.921662764696611e-06,
      "loss": 0.6958,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4061
    },
    {
      "epoch": 0.5523524612455807,
      "grad_norm": 0.61328125,
      "learning_rate": 9.916915159593886e-06,
      "loss": 0.8275,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4062
    },
    {
      "epoch": 0.5524884416644004,
      "grad_norm": 0.349609375,
      "learning_rate": 9.9121675732195e-06,
      "loss": 0.661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4063
    },
    {
      "epoch": 0.55262442208322,
      "grad_norm": 0.2275390625,
      "learning_rate": 9.907420006643619e-06,
      "loss": 0.4111,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4064
    },
    {
      "epoch": 0.5527604025020397,
      "grad_norm": 0.361328125,
      "learning_rate": 9.9026724609364e-06,
      "loss": 0.7183,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4065
    },
    {
      "epoch": 0.5528963829208594,
      "grad_norm": 0.193359375,
      "learning_rate": 9.897924937167996e-06,
      "loss": 0.3841,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4066
    },
    {
      "epoch": 0.5530323633396791,
      "grad_norm": 0.236328125,
      "learning_rate": 9.89317743640856e-06,
      "loss": 0.5156,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4067
    },
    {
      "epoch": 0.5531683437584988,
      "grad_norm": 0.33984375,
      "learning_rate": 9.888429959728236e-06,
      "loss": 0.554,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4068
    },
    {
      "epoch": 0.5533043241773185,
      "grad_norm": 0.75,
      "learning_rate": 9.883682508197163e-06,
      "loss": 0.7323,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4069
    },
    {
      "epoch": 0.5534403045961381,
      "grad_norm": 0.447265625,
      "learning_rate": 9.878935082885476e-06,
      "loss": 0.7046,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4070
    },
    {
      "epoch": 0.5535762850149578,
      "grad_norm": 0.47265625,
      "learning_rate": 9.8741876848633e-06,
      "loss": 0.6901,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4071
    },
    {
      "epoch": 0.5537122654337775,
      "grad_norm": 0.33984375,
      "learning_rate": 9.869440315200758e-06,
      "loss": 0.611,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4072
    },
    {
      "epoch": 0.5538482458525972,
      "grad_norm": 0.28515625,
      "learning_rate": 9.864692974967967e-06,
      "loss": 0.5171,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4073
    },
    {
      "epoch": 0.553984226271417,
      "grad_norm": 0.8046875,
      "learning_rate": 9.859945665235032e-06,
      "loss": 0.577,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4074
    },
    {
      "epoch": 0.5541202066902366,
      "grad_norm": 0.51171875,
      "learning_rate": 9.855198387072052e-06,
      "loss": 0.5692,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4075
    },
    {
      "epoch": 0.5542561871090563,
      "grad_norm": 0.41015625,
      "learning_rate": 9.85045114154913e-06,
      "loss": 0.7863,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4076
    },
    {
      "epoch": 0.554392167527876,
      "grad_norm": 0.388671875,
      "learning_rate": 9.845703929736351e-06,
      "loss": 0.6927,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4077
    },
    {
      "epoch": 0.5545281479466957,
      "grad_norm": 0.4609375,
      "learning_rate": 9.840956752703791e-06,
      "loss": 0.7654,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4078
    },
    {
      "epoch": 0.5546641283655154,
      "grad_norm": 0.5234375,
      "learning_rate": 9.836209611521525e-06,
      "loss": 0.9243,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4079
    },
    {
      "epoch": 0.554800108784335,
      "grad_norm": 0.283203125,
      "learning_rate": 9.831462507259614e-06,
      "loss": 0.5093,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4080
    },
    {
      "epoch": 0.5549360892031547,
      "grad_norm": 0.427734375,
      "learning_rate": 9.826715440988113e-06,
      "loss": 0.5748,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4081
    },
    {
      "epoch": 0.5550720696219744,
      "grad_norm": 0.56640625,
      "learning_rate": 9.821968413777073e-06,
      "loss": 0.5127,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4082
    },
    {
      "epoch": 0.5552080500407941,
      "grad_norm": 0.427734375,
      "learning_rate": 9.817221426696527e-06,
      "loss": 0.674,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4083
    },
    {
      "epoch": 0.5553440304596138,
      "grad_norm": 0.47265625,
      "learning_rate": 9.812474480816508e-06,
      "loss": 0.5775,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4084
    },
    {
      "epoch": 0.5554800108784335,
      "grad_norm": 0.412109375,
      "learning_rate": 9.80772757720703e-06,
      "loss": 0.6809,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4085
    },
    {
      "epoch": 0.5556159912972531,
      "grad_norm": 0.39453125,
      "learning_rate": 9.80298071693811e-06,
      "loss": 0.8293,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4086
    },
    {
      "epoch": 0.5557519717160729,
      "grad_norm": 0.55078125,
      "learning_rate": 9.798233901079745e-06,
      "loss": 0.7503,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4087
    },
    {
      "epoch": 0.5558879521348926,
      "grad_norm": 0.23828125,
      "learning_rate": 9.793487130701925e-06,
      "loss": 0.4471,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4088
    },
    {
      "epoch": 0.5560239325537123,
      "grad_norm": 0.396484375,
      "learning_rate": 9.788740406874632e-06,
      "loss": 0.6509,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4089
    },
    {
      "epoch": 0.556159912972532,
      "grad_norm": 0.423828125,
      "learning_rate": 9.783993730667833e-06,
      "loss": 0.6709,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4090
    },
    {
      "epoch": 0.5562958933913517,
      "grad_norm": 0.6171875,
      "learning_rate": 9.779247103151487e-06,
      "loss": 0.8538,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4091
    },
    {
      "epoch": 0.5564318738101713,
      "grad_norm": 0.341796875,
      "learning_rate": 9.774500525395544e-06,
      "loss": 0.6948,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4092
    },
    {
      "epoch": 0.556567854228991,
      "grad_norm": 0.337890625,
      "learning_rate": 9.76975399846994e-06,
      "loss": 0.4705,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4093
    },
    {
      "epoch": 0.5567038346478107,
      "grad_norm": 0.455078125,
      "learning_rate": 9.765007523444599e-06,
      "loss": 0.6924,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4094
    },
    {
      "epoch": 0.5568398150666304,
      "grad_norm": 0.37890625,
      "learning_rate": 9.760261101389434e-06,
      "loss": 0.7192,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4095
    },
    {
      "epoch": 0.5569757954854501,
      "grad_norm": 0.306640625,
      "learning_rate": 9.75551473337435e-06,
      "loss": 0.613,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4096
    },
    {
      "epoch": 0.5571117759042697,
      "grad_norm": 0.287109375,
      "learning_rate": 9.750768420469233e-06,
      "loss": 0.5566,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4097
    },
    {
      "epoch": 0.5572477563230894,
      "grad_norm": 0.80859375,
      "learning_rate": 9.74602216374396e-06,
      "loss": 0.7288,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4098
    },
    {
      "epoch": 0.5573837367419092,
      "grad_norm": 0.91015625,
      "learning_rate": 9.741275964268402e-06,
      "loss": 0.8944,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4099
    },
    {
      "epoch": 0.5575197171607289,
      "grad_norm": 0.34765625,
      "learning_rate": 9.7365298231124e-06,
      "loss": 0.6315,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4100
    },
    {
      "epoch": 0.5576556975795486,
      "grad_norm": 0.63671875,
      "learning_rate": 9.731783741345798e-06,
      "loss": 0.8652,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4101
    },
    {
      "epoch": 0.5577916779983683,
      "grad_norm": 0.357421875,
      "learning_rate": 9.72703772003842e-06,
      "loss": 0.559,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4102
    },
    {
      "epoch": 0.5579276584171879,
      "grad_norm": 0.5390625,
      "learning_rate": 9.722291760260077e-06,
      "loss": 0.3755,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4103
    },
    {
      "epoch": 0.5580636388360076,
      "grad_norm": 0.33203125,
      "learning_rate": 9.717545863080567e-06,
      "loss": 0.6276,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4104
    },
    {
      "epoch": 0.5581996192548273,
      "grad_norm": 0.26171875,
      "learning_rate": 9.71280002956967e-06,
      "loss": 0.4788,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4105
    },
    {
      "epoch": 0.558335599673647,
      "grad_norm": 0.416015625,
      "learning_rate": 9.708054260797162e-06,
      "loss": 0.6597,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4106
    },
    {
      "epoch": 0.5584715800924667,
      "grad_norm": 0.671875,
      "learning_rate": 9.70330855783279e-06,
      "loss": 0.4637,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4107
    },
    {
      "epoch": 0.5586075605112863,
      "grad_norm": 1.1953125,
      "learning_rate": 9.698562921746298e-06,
      "loss": 0.9836,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4108
    },
    {
      "epoch": 0.558743540930106,
      "grad_norm": 0.384765625,
      "learning_rate": 9.693817353607407e-06,
      "loss": 0.6074,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4109
    },
    {
      "epoch": 0.5588795213489257,
      "grad_norm": 0.412109375,
      "learning_rate": 9.68907185448583e-06,
      "loss": 0.6398,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4110
    },
    {
      "epoch": 0.5590155017677454,
      "grad_norm": 0.5859375,
      "learning_rate": 9.684326425451257e-06,
      "loss": 0.8696,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4111
    },
    {
      "epoch": 0.5591514821865652,
      "grad_norm": 0.26953125,
      "learning_rate": 9.679581067573367e-06,
      "loss": 0.5042,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4112
    },
    {
      "epoch": 0.5592874626053849,
      "grad_norm": 0.6796875,
      "learning_rate": 9.674835781921823e-06,
      "loss": 0.7684,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4113
    },
    {
      "epoch": 0.5594234430242045,
      "grad_norm": 0.69921875,
      "learning_rate": 9.670090569566269e-06,
      "loss": 0.6082,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4114
    },
    {
      "epoch": 0.5595594234430242,
      "grad_norm": 0.451171875,
      "learning_rate": 9.665345431576335e-06,
      "loss": 0.6672,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4115
    },
    {
      "epoch": 0.5596954038618439,
      "grad_norm": 0.63671875,
      "learning_rate": 9.66060036902163e-06,
      "loss": 0.6476,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4116
    },
    {
      "epoch": 0.5598313842806636,
      "grad_norm": 0.75,
      "learning_rate": 9.655855382971751e-06,
      "loss": 0.609,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4117
    },
    {
      "epoch": 0.5599673646994833,
      "grad_norm": 0.40625,
      "learning_rate": 9.651110474496279e-06,
      "loss": 0.7952,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4118
    },
    {
      "epoch": 0.560103345118303,
      "grad_norm": 0.3828125,
      "learning_rate": 9.646365644664771e-06,
      "loss": 0.7482,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4119
    },
    {
      "epoch": 0.5602393255371226,
      "grad_norm": 0.59375,
      "learning_rate": 9.641620894546768e-06,
      "loss": 0.7826,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4120
    },
    {
      "epoch": 0.5603753059559423,
      "grad_norm": 0.32421875,
      "learning_rate": 9.636876225211799e-06,
      "loss": 0.5706,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4121
    },
    {
      "epoch": 0.560511286374762,
      "grad_norm": 0.6015625,
      "learning_rate": 9.632131637729366e-06,
      "loss": 0.7253,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4122
    },
    {
      "epoch": 0.5606472667935817,
      "grad_norm": 0.380859375,
      "learning_rate": 9.627387133168963e-06,
      "loss": 0.6982,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4123
    },
    {
      "epoch": 0.5607832472124014,
      "grad_norm": 0.337890625,
      "learning_rate": 9.622642712600053e-06,
      "loss": 0.6735,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4124
    },
    {
      "epoch": 0.5609192276312212,
      "grad_norm": 0.2451171875,
      "learning_rate": 9.617898377092091e-06,
      "loss": 0.5394,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4125
    },
    {
      "epoch": 0.5610552080500408,
      "grad_norm": 0.423828125,
      "learning_rate": 9.613154127714505e-06,
      "loss": 0.751,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4126
    },
    {
      "epoch": 0.5611911884688605,
      "grad_norm": 0.404296875,
      "learning_rate": 9.60840996553671e-06,
      "loss": 0.7251,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4127
    },
    {
      "epoch": 0.5613271688876802,
      "grad_norm": 0.62890625,
      "learning_rate": 9.603665891628094e-06,
      "loss": 0.5959,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4128
    },
    {
      "epoch": 0.5614631493064999,
      "grad_norm": 0.62890625,
      "learning_rate": 9.598921907058033e-06,
      "loss": 0.6554,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4129
    },
    {
      "epoch": 0.5615991297253196,
      "grad_norm": 0.38671875,
      "learning_rate": 9.594178012895879e-06,
      "loss": 0.4884,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4130
    },
    {
      "epoch": 0.5617351101441392,
      "grad_norm": 0.30078125,
      "learning_rate": 9.58943421021096e-06,
      "loss": 0.5869,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4131
    },
    {
      "epoch": 0.5618710905629589,
      "grad_norm": 0.4921875,
      "learning_rate": 9.584690500072591e-06,
      "loss": 0.5516,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4132
    },
    {
      "epoch": 0.5620070709817786,
      "grad_norm": 0.416015625,
      "learning_rate": 9.579946883550061e-06,
      "loss": 0.6631,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4133
    },
    {
      "epoch": 0.5621430514005983,
      "grad_norm": 0.44921875,
      "learning_rate": 9.575203361712638e-06,
      "loss": 0.6535,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4134
    },
    {
      "epoch": 0.562279031819418,
      "grad_norm": 0.37890625,
      "learning_rate": 9.570459935629567e-06,
      "loss": 0.6991,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4135
    },
    {
      "epoch": 0.5624150122382376,
      "grad_norm": 0.427734375,
      "learning_rate": 9.565716606370083e-06,
      "loss": 0.6258,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4136
    },
    {
      "epoch": 0.5625509926570574,
      "grad_norm": 1.1953125,
      "learning_rate": 9.560973375003385e-06,
      "loss": 0.8779,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4137
    },
    {
      "epoch": 0.5626869730758771,
      "grad_norm": 0.45703125,
      "learning_rate": 9.556230242598656e-06,
      "loss": 0.874,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4138
    },
    {
      "epoch": 0.5628229534946968,
      "grad_norm": 0.310546875,
      "learning_rate": 9.551487210225057e-06,
      "loss": 0.5915,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4139
    },
    {
      "epoch": 0.5629589339135165,
      "grad_norm": 0.5234375,
      "learning_rate": 9.546744278951722e-06,
      "loss": 0.7132,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4140
    },
    {
      "epoch": 0.5630949143323362,
      "grad_norm": 0.30859375,
      "learning_rate": 9.542001449847769e-06,
      "loss": 0.5555,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4141
    },
    {
      "epoch": 0.5632308947511558,
      "grad_norm": 0.5703125,
      "learning_rate": 9.53725872398229e-06,
      "loss": 0.887,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4142
    },
    {
      "epoch": 0.5633668751699755,
      "grad_norm": 0.37109375,
      "learning_rate": 9.53251610242435e-06,
      "loss": 0.7855,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4143
    },
    {
      "epoch": 0.5635028555887952,
      "grad_norm": 0.40234375,
      "learning_rate": 9.527773586242999e-06,
      "loss": 0.4206,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4144
    },
    {
      "epoch": 0.5636388360076149,
      "grad_norm": 0.4296875,
      "learning_rate": 9.523031176507253e-06,
      "loss": 0.6909,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4145
    },
    {
      "epoch": 0.5637748164264346,
      "grad_norm": 0.65625,
      "learning_rate": 9.518288874286111e-06,
      "loss": 0.5381,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4146
    },
    {
      "epoch": 0.5639107968452542,
      "grad_norm": 0.314453125,
      "learning_rate": 9.513546680648546e-06,
      "loss": 0.6392,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4147
    },
    {
      "epoch": 0.5640467772640739,
      "grad_norm": 0.5859375,
      "learning_rate": 9.508804596663509e-06,
      "loss": 0.8521,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4148
    },
    {
      "epoch": 0.5641827576828936,
      "grad_norm": 0.279296875,
      "learning_rate": 9.504062623399919e-06,
      "loss": 0.6694,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4149
    },
    {
      "epoch": 0.5643187381017134,
      "grad_norm": 0.8046875,
      "learning_rate": 9.499320761926679e-06,
      "loss": 0.7124,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4150
    },
    {
      "epoch": 0.5644547185205331,
      "grad_norm": 0.345703125,
      "learning_rate": 9.494579013312657e-06,
      "loss": 0.668,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4151
    },
    {
      "epoch": 0.5645906989393528,
      "grad_norm": 0.283203125,
      "learning_rate": 9.489837378626709e-06,
      "loss": 0.6061,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4152
    },
    {
      "epoch": 0.5647266793581724,
      "grad_norm": 0.375,
      "learning_rate": 9.485095858937653e-06,
      "loss": 0.6621,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4153
    },
    {
      "epoch": 0.5648626597769921,
      "grad_norm": 0.283203125,
      "learning_rate": 9.480354455314283e-06,
      "loss": 0.5352,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4154
    },
    {
      "epoch": 0.5649986401958118,
      "grad_norm": 0.390625,
      "learning_rate": 9.475613168825374e-06,
      "loss": 0.6814,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4155
    },
    {
      "epoch": 0.5651346206146315,
      "grad_norm": 0.64453125,
      "learning_rate": 9.470872000539666e-06,
      "loss": 0.8195,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4156
    },
    {
      "epoch": 0.5652706010334512,
      "grad_norm": 0.318359375,
      "learning_rate": 9.466130951525879e-06,
      "loss": 0.6701,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4157
    },
    {
      "epoch": 0.5654065814522709,
      "grad_norm": 0.4375,
      "learning_rate": 9.461390022852703e-06,
      "loss": 0.6266,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4158
    },
    {
      "epoch": 0.5655425618710905,
      "grad_norm": 0.251953125,
      "learning_rate": 9.4566492155888e-06,
      "loss": 0.5247,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4159
    },
    {
      "epoch": 0.5656785422899102,
      "grad_norm": 0.392578125,
      "learning_rate": 9.451908530802809e-06,
      "loss": 0.7336,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4160
    },
    {
      "epoch": 0.5658145227087299,
      "grad_norm": 7.15625,
      "learning_rate": 9.447167969563334e-06,
      "loss": 0.8049,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4161
    },
    {
      "epoch": 0.5659505031275497,
      "grad_norm": 0.45703125,
      "learning_rate": 9.442427532938958e-06,
      "loss": 0.4959,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4162
    },
    {
      "epoch": 0.5660864835463694,
      "grad_norm": 0.36328125,
      "learning_rate": 9.437687221998232e-06,
      "loss": 0.7135,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4163
    },
    {
      "epoch": 0.566222463965189,
      "grad_norm": 0.408203125,
      "learning_rate": 9.432947037809683e-06,
      "loss": 0.8081,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4164
    },
    {
      "epoch": 0.5663584443840087,
      "grad_norm": 0.9765625,
      "learning_rate": 9.428206981441804e-06,
      "loss": 0.6777,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4165
    },
    {
      "epoch": 0.5664944248028284,
      "grad_norm": 0.46484375,
      "learning_rate": 9.423467053963063e-06,
      "loss": 0.7674,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4166
    },
    {
      "epoch": 0.5666304052216481,
      "grad_norm": 0.390625,
      "learning_rate": 9.418727256441895e-06,
      "loss": 0.5936,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4167
    },
    {
      "epoch": 0.5667663856404678,
      "grad_norm": 0.291015625,
      "learning_rate": 9.41398758994671e-06,
      "loss": 0.4957,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4168
    },
    {
      "epoch": 0.5669023660592875,
      "grad_norm": 0.357421875,
      "learning_rate": 9.409248055545889e-06,
      "loss": 0.7266,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4169
    },
    {
      "epoch": 0.5670383464781071,
      "grad_norm": 0.330078125,
      "learning_rate": 9.404508654307778e-06,
      "loss": 0.4956,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4170
    },
    {
      "epoch": 0.5671743268969268,
      "grad_norm": 0.37109375,
      "learning_rate": 9.399769387300697e-06,
      "loss": 0.4767,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4171
    },
    {
      "epoch": 0.5673103073157465,
      "grad_norm": 0.423828125,
      "learning_rate": 9.395030255592933e-06,
      "loss": 0.5936,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4172
    },
    {
      "epoch": 0.5674462877345662,
      "grad_norm": 0.32421875,
      "learning_rate": 9.390291260252748e-06,
      "loss": 0.4805,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4173
    },
    {
      "epoch": 0.5675822681533859,
      "grad_norm": 0.2890625,
      "learning_rate": 9.385552402348364e-06,
      "loss": 0.5845,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4174
    },
    {
      "epoch": 0.5677182485722057,
      "grad_norm": 0.29296875,
      "learning_rate": 9.380813682947985e-06,
      "loss": 0.5545,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4175
    },
    {
      "epoch": 0.5678542289910253,
      "grad_norm": 0.45703125,
      "learning_rate": 9.37607510311977e-06,
      "loss": 0.5936,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4176
    },
    {
      "epoch": 0.567990209409845,
      "grad_norm": 0.6640625,
      "learning_rate": 9.371336663931854e-06,
      "loss": 0.6735,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4177
    },
    {
      "epoch": 0.5681261898286647,
      "grad_norm": 0.734375,
      "learning_rate": 9.36659836645234e-06,
      "loss": 0.8677,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4178
    },
    {
      "epoch": 0.5682621702474844,
      "grad_norm": 0.4609375,
      "learning_rate": 9.361860211749299e-06,
      "loss": 0.7718,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4179
    },
    {
      "epoch": 0.5683981506663041,
      "grad_norm": 0.267578125,
      "learning_rate": 9.35712220089077e-06,
      "loss": 0.5239,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4180
    },
    {
      "epoch": 0.5685341310851237,
      "grad_norm": 0.333984375,
      "learning_rate": 9.352384334944754e-06,
      "loss": 0.5492,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4181
    },
    {
      "epoch": 0.5686701115039434,
      "grad_norm": 0.5390625,
      "learning_rate": 9.34764661497923e-06,
      "loss": 0.7573,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4182
    },
    {
      "epoch": 0.5688060919227631,
      "grad_norm": 0.9296875,
      "learning_rate": 9.342909042062131e-06,
      "loss": 0.7091,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4183
    },
    {
      "epoch": 0.5689420723415828,
      "grad_norm": 0.451171875,
      "learning_rate": 9.33817161726137e-06,
      "loss": 0.6722,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4184
    },
    {
      "epoch": 0.5690780527604025,
      "grad_norm": 0.4375,
      "learning_rate": 9.333434341644816e-06,
      "loss": 0.7166,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4185
    },
    {
      "epoch": 0.5692140331792221,
      "grad_norm": 0.4921875,
      "learning_rate": 9.328697216280313e-06,
      "loss": 0.6663,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4186
    },
    {
      "epoch": 0.5693500135980419,
      "grad_norm": 0.546875,
      "learning_rate": 9.323960242235663e-06,
      "loss": 0.6644,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4187
    },
    {
      "epoch": 0.5694859940168616,
      "grad_norm": 0.35546875,
      "learning_rate": 9.319223420578642e-06,
      "loss": 0.7788,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4188
    },
    {
      "epoch": 0.5696219744356813,
      "grad_norm": 5.09375,
      "learning_rate": 9.314486752376984e-06,
      "loss": 0.7094,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4189
    },
    {
      "epoch": 0.569757954854501,
      "grad_norm": 0.4140625,
      "learning_rate": 9.309750238698393e-06,
      "loss": 0.7199,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4190
    },
    {
      "epoch": 0.5698939352733207,
      "grad_norm": 0.50390625,
      "learning_rate": 9.305013880610536e-06,
      "loss": 0.797,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4191
    },
    {
      "epoch": 0.5700299156921403,
      "grad_norm": 0.33984375,
      "learning_rate": 9.30027767918105e-06,
      "loss": 0.5021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4192
    },
    {
      "epoch": 0.57016589611096,
      "grad_norm": 0.3359375,
      "learning_rate": 9.295541635477527e-06,
      "loss": 0.6388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4193
    },
    {
      "epoch": 0.5703018765297797,
      "grad_norm": 0.52734375,
      "learning_rate": 9.290805750567532e-06,
      "loss": 0.8442,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4194
    },
    {
      "epoch": 0.5704378569485994,
      "grad_norm": 0.73046875,
      "learning_rate": 9.286070025518589e-06,
      "loss": 0.6255,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4195
    },
    {
      "epoch": 0.5705738373674191,
      "grad_norm": 0.44921875,
      "learning_rate": 9.281334461398196e-06,
      "loss": 0.6479,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4196
    },
    {
      "epoch": 0.5707098177862387,
      "grad_norm": 0.384765625,
      "learning_rate": 9.276599059273801e-06,
      "loss": 0.6291,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4197
    },
    {
      "epoch": 0.5708457982050584,
      "grad_norm": 0.2314453125,
      "learning_rate": 9.271863820212822e-06,
      "loss": 0.3545,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4198
    },
    {
      "epoch": 0.5709817786238781,
      "grad_norm": 0.68359375,
      "learning_rate": 9.267128745282642e-06,
      "loss": 0.5549,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4199
    },
    {
      "epoch": 0.5711177590426979,
      "grad_norm": 0.28515625,
      "learning_rate": 9.262393835550601e-06,
      "loss": 0.5024,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4200
    },
    {
      "epoch": 0.5712537394615176,
      "grad_norm": 0.310546875,
      "learning_rate": 9.257659092084007e-06,
      "loss": 0.5586,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4201
    },
    {
      "epoch": 0.5713897198803373,
      "grad_norm": 0.423828125,
      "learning_rate": 9.25292451595013e-06,
      "loss": 0.7,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4202
    },
    {
      "epoch": 0.571525700299157,
      "grad_norm": 0.390625,
      "learning_rate": 9.248190108216203e-06,
      "loss": 0.8376,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4203
    },
    {
      "epoch": 0.5716616807179766,
      "grad_norm": 0.32421875,
      "learning_rate": 9.243455869949414e-06,
      "loss": 0.6156,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4204
    },
    {
      "epoch": 0.5717976611367963,
      "grad_norm": 0.490234375,
      "learning_rate": 9.238721802216923e-06,
      "loss": 0.6387,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4205
    },
    {
      "epoch": 0.571933641555616,
      "grad_norm": 0.38671875,
      "learning_rate": 9.233987906085844e-06,
      "loss": 0.5303,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4206
    },
    {
      "epoch": 0.5720696219744357,
      "grad_norm": 0.361328125,
      "learning_rate": 9.22925418262326e-06,
      "loss": 0.5913,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4207
    },
    {
      "epoch": 0.5722056023932554,
      "grad_norm": 0.341796875,
      "learning_rate": 9.224520632896206e-06,
      "loss": 0.5505,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4208
    },
    {
      "epoch": 0.572341582812075,
      "grad_norm": 0.51953125,
      "learning_rate": 9.219787257971679e-06,
      "loss": 0.436,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4209
    },
    {
      "epoch": 0.5724775632308947,
      "grad_norm": 0.32421875,
      "learning_rate": 9.215054058916646e-06,
      "loss": 0.5641,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4210
    },
    {
      "epoch": 0.5726135436497144,
      "grad_norm": 0.37890625,
      "learning_rate": 9.210321036798023e-06,
      "loss": 0.6438,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4211
    },
    {
      "epoch": 0.5727495240685341,
      "grad_norm": 1.6171875,
      "learning_rate": 9.205588192682693e-06,
      "loss": 0.5544,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4212
    },
    {
      "epoch": 0.5728855044873539,
      "grad_norm": 0.482421875,
      "learning_rate": 9.200855527637496e-06,
      "loss": 0.8208,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4213
    },
    {
      "epoch": 0.5730214849061736,
      "grad_norm": 1.90625,
      "learning_rate": 9.196123042729236e-06,
      "loss": 0.8584,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4214
    },
    {
      "epoch": 0.5731574653249932,
      "grad_norm": 0.330078125,
      "learning_rate": 9.19139073902467e-06,
      "loss": 0.5225,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4215
    },
    {
      "epoch": 0.5732934457438129,
      "grad_norm": 0.4296875,
      "learning_rate": 9.186658617590516e-06,
      "loss": 0.6891,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4216
    },
    {
      "epoch": 0.5734294261626326,
      "grad_norm": 0.388671875,
      "learning_rate": 9.181926679493454e-06,
      "loss": 0.6756,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4217
    },
    {
      "epoch": 0.5735654065814523,
      "grad_norm": 0.361328125,
      "learning_rate": 9.177194925800121e-06,
      "loss": 0.5354,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4218
    },
    {
      "epoch": 0.573701387000272,
      "grad_norm": 0.361328125,
      "learning_rate": 9.172463357577111e-06,
      "loss": 0.5697,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4219
    },
    {
      "epoch": 0.5738373674190916,
      "grad_norm": 0.54296875,
      "learning_rate": 9.167731975890977e-06,
      "loss": 0.6506,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4220
    },
    {
      "epoch": 0.5739733478379113,
      "grad_norm": 0.28515625,
      "learning_rate": 9.163000781808232e-06,
      "loss": 0.6061,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4221
    },
    {
      "epoch": 0.574109328256731,
      "grad_norm": 0.87109375,
      "learning_rate": 9.158269776395343e-06,
      "loss": 0.6868,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4222
    },
    {
      "epoch": 0.5742453086755507,
      "grad_norm": 0.404296875,
      "learning_rate": 9.153538960718738e-06,
      "loss": 0.818,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4223
    },
    {
      "epoch": 0.5743812890943704,
      "grad_norm": 0.35546875,
      "learning_rate": 9.1488083358448e-06,
      "loss": 0.5785,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4224
    },
    {
      "epoch": 0.5745172695131902,
      "grad_norm": 0.3125,
      "learning_rate": 9.14407790283987e-06,
      "loss": 0.5957,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4225
    },
    {
      "epoch": 0.5746532499320098,
      "grad_norm": 0.349609375,
      "learning_rate": 9.139347662770248e-06,
      "loss": 0.6675,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4226
    },
    {
      "epoch": 0.5747892303508295,
      "grad_norm": 0.3046875,
      "learning_rate": 9.134617616702182e-06,
      "loss": 0.5786,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4227
    },
    {
      "epoch": 0.5749252107696492,
      "grad_norm": 0.30078125,
      "learning_rate": 9.129887765701885e-06,
      "loss": 0.3937,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4228
    },
    {
      "epoch": 0.5750611911884689,
      "grad_norm": 0.31640625,
      "learning_rate": 9.125158110835525e-06,
      "loss": 0.7004,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4229
    },
    {
      "epoch": 0.5751971716072886,
      "grad_norm": 0.40625,
      "learning_rate": 9.120428653169222e-06,
      "loss": 0.812,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4230
    },
    {
      "epoch": 0.5753331520261082,
      "grad_norm": 0.48828125,
      "learning_rate": 9.115699393769055e-06,
      "loss": 0.6243,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4231
    },
    {
      "epoch": 0.5754691324449279,
      "grad_norm": 0.28515625,
      "learning_rate": 9.110970333701057e-06,
      "loss": 0.5602,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4232
    },
    {
      "epoch": 0.5756051128637476,
      "grad_norm": 0.2197265625,
      "learning_rate": 9.106241474031213e-06,
      "loss": 0.4012,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4233
    },
    {
      "epoch": 0.5757410932825673,
      "grad_norm": 0.30078125,
      "learning_rate": 9.101512815825469e-06,
      "loss": 0.5374,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4234
    },
    {
      "epoch": 0.575877073701387,
      "grad_norm": 0.375,
      "learning_rate": 9.096784360149722e-06,
      "loss": 0.588,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4235
    },
    {
      "epoch": 0.5760130541202066,
      "grad_norm": 0.34765625,
      "learning_rate": 9.092056108069822e-06,
      "loss": 0.6406,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4236
    },
    {
      "epoch": 0.5761490345390263,
      "grad_norm": 0.51953125,
      "learning_rate": 9.087328060651576e-06,
      "loss": 0.6392,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4237
    },
    {
      "epoch": 0.5762850149578461,
      "grad_norm": 0.40625,
      "learning_rate": 9.082600218960743e-06,
      "loss": 0.7349,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4238
    },
    {
      "epoch": 0.5764209953766658,
      "grad_norm": 0.30859375,
      "learning_rate": 9.077872584063037e-06,
      "loss": 0.6543,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4239
    },
    {
      "epoch": 0.5765569757954855,
      "grad_norm": 0.515625,
      "learning_rate": 9.073145157024126e-06,
      "loss": 0.784,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4240
    },
    {
      "epoch": 0.5766929562143052,
      "grad_norm": 0.5703125,
      "learning_rate": 9.068417938909628e-06,
      "loss": 0.695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4241
    },
    {
      "epoch": 0.5768289366331248,
      "grad_norm": 0.365234375,
      "learning_rate": 9.063690930785116e-06,
      "loss": 0.7038,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4242
    },
    {
      "epoch": 0.5769649170519445,
      "grad_norm": 0.345703125,
      "learning_rate": 9.058964133716113e-06,
      "loss": 0.5802,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4243
    },
    {
      "epoch": 0.5771008974707642,
      "grad_norm": 0.486328125,
      "learning_rate": 9.054237548768103e-06,
      "loss": 0.6692,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4244
    },
    {
      "epoch": 0.5772368778895839,
      "grad_norm": 0.439453125,
      "learning_rate": 9.049511177006512e-06,
      "loss": 0.5453,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4245
    },
    {
      "epoch": 0.5773728583084036,
      "grad_norm": 0.41015625,
      "learning_rate": 9.04478501949672e-06,
      "loss": 0.6683,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4246
    },
    {
      "epoch": 0.5775088387272233,
      "grad_norm": 0.44921875,
      "learning_rate": 9.040059077304065e-06,
      "loss": 0.7368,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4247
    },
    {
      "epoch": 0.5776448191460429,
      "grad_norm": 1.65625,
      "learning_rate": 9.03533335149383e-06,
      "loss": 0.7036,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4248
    },
    {
      "epoch": 0.5777807995648626,
      "grad_norm": 0.703125,
      "learning_rate": 9.030607843131251e-06,
      "loss": 0.7962,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4249
    },
    {
      "epoch": 0.5779167799836824,
      "grad_norm": 0.30859375,
      "learning_rate": 9.025882553281518e-06,
      "loss": 0.6872,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4250
    },
    {
      "epoch": 0.5780527604025021,
      "grad_norm": 0.6640625,
      "learning_rate": 9.021157483009766e-06,
      "loss": 0.8265,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4251
    },
    {
      "epoch": 0.5781887408213218,
      "grad_norm": 0.306640625,
      "learning_rate": 9.016432633381083e-06,
      "loss": 0.5815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4252
    },
    {
      "epoch": 0.5783247212401414,
      "grad_norm": 0.443359375,
      "learning_rate": 9.011708005460512e-06,
      "loss": 0.6868,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4253
    },
    {
      "epoch": 0.5784607016589611,
      "grad_norm": 0.72265625,
      "learning_rate": 9.006983600313038e-06,
      "loss": 0.7259,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4254
    },
    {
      "epoch": 0.5785966820777808,
      "grad_norm": 0.291015625,
      "learning_rate": 9.002259419003602e-06,
      "loss": 0.715,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4255
    },
    {
      "epoch": 0.5787326624966005,
      "grad_norm": 0.60546875,
      "learning_rate": 8.997535462597087e-06,
      "loss": 0.533,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4256
    },
    {
      "epoch": 0.5788686429154202,
      "grad_norm": 0.40234375,
      "learning_rate": 8.99281173215834e-06,
      "loss": 0.6929,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4257
    },
    {
      "epoch": 0.5790046233342399,
      "grad_norm": 0.298828125,
      "learning_rate": 8.98808822875214e-06,
      "loss": 0.6243,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4258
    },
    {
      "epoch": 0.5791406037530595,
      "grad_norm": 0.69921875,
      "learning_rate": 8.983364953443227e-06,
      "loss": 0.7183,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4259
    },
    {
      "epoch": 0.5792765841718792,
      "grad_norm": 0.423828125,
      "learning_rate": 8.978641907296283e-06,
      "loss": 0.4939,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4260
    },
    {
      "epoch": 0.5794125645906989,
      "grad_norm": 0.314453125,
      "learning_rate": 8.973919091375942e-06,
      "loss": 0.6141,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4261
    },
    {
      "epoch": 0.5795485450095186,
      "grad_norm": 0.376953125,
      "learning_rate": 8.969196506746781e-06,
      "loss": 0.2116,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4262
    },
    {
      "epoch": 0.5796845254283384,
      "grad_norm": 0.42578125,
      "learning_rate": 8.96447415447333e-06,
      "loss": 0.7725,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4263
    },
    {
      "epoch": 0.579820505847158,
      "grad_norm": 0.38671875,
      "learning_rate": 8.959752035620065e-06,
      "loss": 0.7773,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4264
    },
    {
      "epoch": 0.5799564862659777,
      "grad_norm": 0.359375,
      "learning_rate": 8.955030151251409e-06,
      "loss": 0.6966,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4265
    },
    {
      "epoch": 0.5800924666847974,
      "grad_norm": 0.326171875,
      "learning_rate": 8.950308502431733e-06,
      "loss": 0.6108,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4266
    },
    {
      "epoch": 0.5802284471036171,
      "grad_norm": 3.796875,
      "learning_rate": 8.945587090225354e-06,
      "loss": 0.8626,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4267
    },
    {
      "epoch": 0.5803644275224368,
      "grad_norm": 0.447265625,
      "learning_rate": 8.940865915696537e-06,
      "loss": 0.6172,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4268
    },
    {
      "epoch": 0.5805004079412565,
      "grad_norm": 0.62109375,
      "learning_rate": 8.93614497990949e-06,
      "loss": 0.5013,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4269
    },
    {
      "epoch": 0.5806363883600761,
      "grad_norm": 0.40234375,
      "learning_rate": 8.93142428392837e-06,
      "loss": 0.777,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4270
    },
    {
      "epoch": 0.5807723687788958,
      "grad_norm": 1.21875,
      "learning_rate": 8.926703828817285e-06,
      "loss": 0.7124,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4271
    },
    {
      "epoch": 0.5809083491977155,
      "grad_norm": 0.330078125,
      "learning_rate": 8.921983615640277e-06,
      "loss": 0.6639,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4272
    },
    {
      "epoch": 0.5810443296165352,
      "grad_norm": 0.68359375,
      "learning_rate": 8.917263645461343e-06,
      "loss": 0.738,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4273
    },
    {
      "epoch": 0.5811803100353549,
      "grad_norm": 0.365234375,
      "learning_rate": 8.91254391934442e-06,
      "loss": 0.5991,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4274
    },
    {
      "epoch": 0.5813162904541745,
      "grad_norm": 0.625,
      "learning_rate": 8.907824438353393e-06,
      "loss": 0.5526,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4275
    },
    {
      "epoch": 0.5814522708729943,
      "grad_norm": 0.408203125,
      "learning_rate": 8.90310520355209e-06,
      "loss": 0.9068,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4276
    },
    {
      "epoch": 0.581588251291814,
      "grad_norm": 0.466796875,
      "learning_rate": 8.898386216004287e-06,
      "loss": 0.8197,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4277
    },
    {
      "epoch": 0.5817242317106337,
      "grad_norm": 0.380859375,
      "learning_rate": 8.893667476773698e-06,
      "loss": 0.7713,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4278
    },
    {
      "epoch": 0.5818602121294534,
      "grad_norm": 0.384765625,
      "learning_rate": 8.888948986923984e-06,
      "loss": 0.5668,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4279
    },
    {
      "epoch": 0.5819961925482731,
      "grad_norm": 0.5625,
      "learning_rate": 8.884230747518755e-06,
      "loss": 0.8693,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4280
    },
    {
      "epoch": 0.5821321729670927,
      "grad_norm": 0.421875,
      "learning_rate": 8.879512759621556e-06,
      "loss": 0.6628,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4281
    },
    {
      "epoch": 0.5822681533859124,
      "grad_norm": 0.3046875,
      "learning_rate": 8.87479502429588e-06,
      "loss": 0.5501,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4282
    },
    {
      "epoch": 0.5824041338047321,
      "grad_norm": 0.484375,
      "learning_rate": 8.870077542605163e-06,
      "loss": 0.8542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4283
    },
    {
      "epoch": 0.5825401142235518,
      "grad_norm": 0.34375,
      "learning_rate": 8.865360315612781e-06,
      "loss": 0.6047,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4284
    },
    {
      "epoch": 0.5826760946423715,
      "grad_norm": 0.59375,
      "learning_rate": 8.860643344382057e-06,
      "loss": 0.5851,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4285
    },
    {
      "epoch": 0.5828120750611911,
      "grad_norm": 0.265625,
      "learning_rate": 8.855926629976252e-06,
      "loss": 0.5215,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4286
    },
    {
      "epoch": 0.5829480554800108,
      "grad_norm": 0.38671875,
      "learning_rate": 8.851210173458573e-06,
      "loss": 0.743,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4287
    },
    {
      "epoch": 0.5830840358988306,
      "grad_norm": 0.376953125,
      "learning_rate": 8.846493975892166e-06,
      "loss": 0.5718,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4288
    },
    {
      "epoch": 0.5832200163176503,
      "grad_norm": 0.421875,
      "learning_rate": 8.84177803834012e-06,
      "loss": 0.6702,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4289
    },
    {
      "epoch": 0.58335599673647,
      "grad_norm": 0.55078125,
      "learning_rate": 8.837062361865465e-06,
      "loss": 0.7555,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4290
    },
    {
      "epoch": 0.5834919771552897,
      "grad_norm": 0.35546875,
      "learning_rate": 8.832346947531171e-06,
      "loss": 0.4993,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4291
    },
    {
      "epoch": 0.5836279575741093,
      "grad_norm": 0.51171875,
      "learning_rate": 8.827631796400152e-06,
      "loss": 0.6738,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4292
    },
    {
      "epoch": 0.583763937992929,
      "grad_norm": 0.54296875,
      "learning_rate": 8.822916909535258e-06,
      "loss": 0.8556,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4293
    },
    {
      "epoch": 0.5838999184117487,
      "grad_norm": 0.40625,
      "learning_rate": 8.818202287999285e-06,
      "loss": 0.6657,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4294
    },
    {
      "epoch": 0.5840358988305684,
      "grad_norm": 0.39453125,
      "learning_rate": 8.813487932854963e-06,
      "loss": 0.4635,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4295
    },
    {
      "epoch": 0.5841718792493881,
      "grad_norm": 0.384765625,
      "learning_rate": 8.808773845164969e-06,
      "loss": 0.6849,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4296
    },
    {
      "epoch": 0.5843078596682078,
      "grad_norm": 0.470703125,
      "learning_rate": 8.804060025991914e-06,
      "loss": 0.6341,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4297
    },
    {
      "epoch": 0.5844438400870274,
      "grad_norm": 0.3203125,
      "learning_rate": 8.799346476398351e-06,
      "loss": 0.6717,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4298
    },
    {
      "epoch": 0.5845798205058471,
      "grad_norm": 0.54296875,
      "learning_rate": 8.79463319744677e-06,
      "loss": 0.5046,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4299
    },
    {
      "epoch": 0.5847158009246668,
      "grad_norm": 0.38671875,
      "learning_rate": 8.789920190199607e-06,
      "loss": 0.6584,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4300
    },
    {
      "epoch": 0.5848517813434866,
      "grad_norm": 0.3984375,
      "learning_rate": 8.785207455719224e-06,
      "loss": 0.6838,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4301
    },
    {
      "epoch": 0.5849877617623063,
      "grad_norm": 0.263671875,
      "learning_rate": 8.780494995067934e-06,
      "loss": 0.5404,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4302
    },
    {
      "epoch": 0.585123742181126,
      "grad_norm": 0.244140625,
      "learning_rate": 8.775782809307983e-06,
      "loss": 0.564,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4303
    },
    {
      "epoch": 0.5852597225999456,
      "grad_norm": 1.109375,
      "learning_rate": 8.771070899501552e-06,
      "loss": 0.7622,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4304
    },
    {
      "epoch": 0.5853957030187653,
      "grad_norm": 0.291015625,
      "learning_rate": 8.766359266710767e-06,
      "loss": 0.5247,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4305
    },
    {
      "epoch": 0.585531683437585,
      "grad_norm": 0.341796875,
      "learning_rate": 8.761647911997684e-06,
      "loss": 0.6003,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4306
    },
    {
      "epoch": 0.5856676638564047,
      "grad_norm": 0.330078125,
      "learning_rate": 8.756936836424303e-06,
      "loss": 0.633,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4307
    },
    {
      "epoch": 0.5858036442752244,
      "grad_norm": 0.52734375,
      "learning_rate": 8.752226041052556e-06,
      "loss": 0.6021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4308
    },
    {
      "epoch": 0.585939624694044,
      "grad_norm": 0.44921875,
      "learning_rate": 8.747515526944318e-06,
      "loss": 0.7816,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4309
    },
    {
      "epoch": 0.5860756051128637,
      "grad_norm": 0.328125,
      "learning_rate": 8.74280529516139e-06,
      "loss": 0.7419,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4310
    },
    {
      "epoch": 0.5862115855316834,
      "grad_norm": 0.5703125,
      "learning_rate": 8.738095346765519e-06,
      "loss": 0.8221,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4311
    },
    {
      "epoch": 0.5863475659505031,
      "grad_norm": 0.33203125,
      "learning_rate": 8.733385682818384e-06,
      "loss": 0.7314,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4312
    },
    {
      "epoch": 0.5864835463693229,
      "grad_norm": 0.4609375,
      "learning_rate": 8.728676304381604e-06,
      "loss": 0.7769,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4313
    },
    {
      "epoch": 0.5866195267881426,
      "grad_norm": 0.51171875,
      "learning_rate": 8.723967212516729e-06,
      "loss": 0.7122,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4314
    },
    {
      "epoch": 0.5867555072069622,
      "grad_norm": 0.34765625,
      "learning_rate": 8.719258408285244e-06,
      "loss": 0.6463,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4315
    },
    {
      "epoch": 0.5868914876257819,
      "grad_norm": 0.416015625,
      "learning_rate": 8.714549892748569e-06,
      "loss": 0.7202,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4316
    },
    {
      "epoch": 0.5870274680446016,
      "grad_norm": 0.51171875,
      "learning_rate": 8.709841666968069e-06,
      "loss": 0.7687,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4317
    },
    {
      "epoch": 0.5871634484634213,
      "grad_norm": 0.498046875,
      "learning_rate": 8.705133732005029e-06,
      "loss": 0.8013,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4318
    },
    {
      "epoch": 0.587299428882241,
      "grad_norm": 0.4609375,
      "learning_rate": 8.700426088920677e-06,
      "loss": 0.8218,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4319
    },
    {
      "epoch": 0.5874354093010606,
      "grad_norm": 0.33984375,
      "learning_rate": 8.695718738776173e-06,
      "loss": 0.706,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4320
    },
    {
      "epoch": 0.5875713897198803,
      "grad_norm": 0.421875,
      "learning_rate": 8.691011682632613e-06,
      "loss": 0.6822,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4321
    },
    {
      "epoch": 0.5877073701387,
      "grad_norm": 0.318359375,
      "learning_rate": 8.686304921551021e-06,
      "loss": 0.6807,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4322
    },
    {
      "epoch": 0.5878433505575197,
      "grad_norm": 0.37890625,
      "learning_rate": 8.681598456592363e-06,
      "loss": 0.6712,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4323
    },
    {
      "epoch": 0.5879793309763394,
      "grad_norm": 0.3359375,
      "learning_rate": 8.676892288817531e-06,
      "loss": 0.6652,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4324
    },
    {
      "epoch": 0.588115311395159,
      "grad_norm": 0.66796875,
      "learning_rate": 8.672186419287353e-06,
      "loss": 0.7925,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4325
    },
    {
      "epoch": 0.5882512918139788,
      "grad_norm": 0.5078125,
      "learning_rate": 8.667480849062592e-06,
      "loss": 0.5887,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4326
    },
    {
      "epoch": 0.5883872722327985,
      "grad_norm": 0.84375,
      "learning_rate": 8.662775579203937e-06,
      "loss": 0.8407,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4327
    },
    {
      "epoch": 0.5885232526516182,
      "grad_norm": 0.34375,
      "learning_rate": 8.658070610772016e-06,
      "loss": 0.7215,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4328
    },
    {
      "epoch": 0.5886592330704379,
      "grad_norm": 0.69140625,
      "learning_rate": 8.653365944827387e-06,
      "loss": 0.751,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4329
    },
    {
      "epoch": 0.5887952134892576,
      "grad_norm": 0.3515625,
      "learning_rate": 8.648661582430538e-06,
      "loss": 0.7306,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4330
    },
    {
      "epoch": 0.5889311939080772,
      "grad_norm": 0.4140625,
      "learning_rate": 8.643957524641889e-06,
      "loss": 0.7249,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4331
    },
    {
      "epoch": 0.5890671743268969,
      "grad_norm": 0.3984375,
      "learning_rate": 8.639253772521794e-06,
      "loss": 0.4883,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4332
    },
    {
      "epoch": 0.5892031547457166,
      "grad_norm": 0.376953125,
      "learning_rate": 8.634550327130533e-06,
      "loss": 0.7738,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4333
    },
    {
      "epoch": 0.5893391351645363,
      "grad_norm": 0.3671875,
      "learning_rate": 8.629847189528325e-06,
      "loss": 0.5889,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4334
    },
    {
      "epoch": 0.589475115583356,
      "grad_norm": 0.322265625,
      "learning_rate": 8.62514436077531e-06,
      "loss": 0.5895,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4335
    },
    {
      "epoch": 0.5896110960021756,
      "grad_norm": 0.2431640625,
      "learning_rate": 8.620441841931566e-06,
      "loss": 0.459,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4336
    },
    {
      "epoch": 0.5897470764209953,
      "grad_norm": 0.91015625,
      "learning_rate": 8.615739634057098e-06,
      "loss": 0.6761,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4337
    },
    {
      "epoch": 0.589883056839815,
      "grad_norm": 0.279296875,
      "learning_rate": 8.61103773821184e-06,
      "loss": 0.5949,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4338
    },
    {
      "epoch": 0.5900190372586348,
      "grad_norm": 0.447265625,
      "learning_rate": 8.606336155455658e-06,
      "loss": 0.8014,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4339
    },
    {
      "epoch": 0.5901550176774545,
      "grad_norm": 0.341796875,
      "learning_rate": 8.601634886848345e-06,
      "loss": 0.5387,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4340
    },
    {
      "epoch": 0.5902909980962742,
      "grad_norm": 0.3671875,
      "learning_rate": 8.596933933449625e-06,
      "loss": 0.6478,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4341
    },
    {
      "epoch": 0.5904269785150938,
      "grad_norm": 0.71875,
      "learning_rate": 8.59223329631915e-06,
      "loss": 0.7043,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4342
    },
    {
      "epoch": 0.5905629589339135,
      "grad_norm": 0.62890625,
      "learning_rate": 8.587532976516503e-06,
      "loss": 0.6181,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4343
    },
    {
      "epoch": 0.5906989393527332,
      "grad_norm": 0.28515625,
      "learning_rate": 8.582832975101191e-06,
      "loss": 0.5757,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4344
    },
    {
      "epoch": 0.5908349197715529,
      "grad_norm": 0.380859375,
      "learning_rate": 8.578133293132652e-06,
      "loss": 0.7505,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4345
    },
    {
      "epoch": 0.5909709001903726,
      "grad_norm": 0.45703125,
      "learning_rate": 8.573433931670255e-06,
      "loss": 0.8218,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4346
    },
    {
      "epoch": 0.5911068806091923,
      "grad_norm": 0.34765625,
      "learning_rate": 8.568734891773288e-06,
      "loss": 0.7186,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4347
    },
    {
      "epoch": 0.5912428610280119,
      "grad_norm": 0.5859375,
      "learning_rate": 8.564036174500977e-06,
      "loss": 0.8208,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4348
    },
    {
      "epoch": 0.5913788414468316,
      "grad_norm": 0.3203125,
      "learning_rate": 8.559337780912469e-06,
      "loss": 0.5183,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4349
    },
    {
      "epoch": 0.5915148218656513,
      "grad_norm": 10.5,
      "learning_rate": 8.554639712066837e-06,
      "loss": 0.8151,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4350
    },
    {
      "epoch": 0.5916508022844711,
      "grad_norm": 0.462890625,
      "learning_rate": 8.549941969023086e-06,
      "loss": 0.5962,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4351
    },
    {
      "epoch": 0.5917867827032908,
      "grad_norm": 1.0859375,
      "learning_rate": 8.545244552840144e-06,
      "loss": 0.6079,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4352
    },
    {
      "epoch": 0.5919227631221105,
      "grad_norm": 0.3359375,
      "learning_rate": 8.540547464576864e-06,
      "loss": 0.6313,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4353
    },
    {
      "epoch": 0.5920587435409301,
      "grad_norm": 0.294921875,
      "learning_rate": 8.535850705292032e-06,
      "loss": 0.521,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4354
    },
    {
      "epoch": 0.5921947239597498,
      "grad_norm": 0.326171875,
      "learning_rate": 8.53115427604435e-06,
      "loss": 0.5423,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4355
    },
    {
      "epoch": 0.5923307043785695,
      "grad_norm": 0.34375,
      "learning_rate": 8.526458177892452e-06,
      "loss": 0.464,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4356
    },
    {
      "epoch": 0.5924666847973892,
      "grad_norm": 0.2490234375,
      "learning_rate": 8.521762411894897e-06,
      "loss": 0.4318,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4357
    },
    {
      "epoch": 0.5926026652162089,
      "grad_norm": 0.443359375,
      "learning_rate": 8.517066979110168e-06,
      "loss": 0.5768,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4358
    },
    {
      "epoch": 0.5927386456350285,
      "grad_norm": 0.388671875,
      "learning_rate": 8.51237188059667e-06,
      "loss": 0.7653,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4359
    },
    {
      "epoch": 0.5928746260538482,
      "grad_norm": 0.3359375,
      "learning_rate": 8.507677117412741e-06,
      "loss": 0.6343,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4360
    },
    {
      "epoch": 0.5930106064726679,
      "grad_norm": 0.349609375,
      "learning_rate": 8.502982690616633e-06,
      "loss": 0.6672,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4361
    },
    {
      "epoch": 0.5931465868914876,
      "grad_norm": 0.66796875,
      "learning_rate": 8.49828860126653e-06,
      "loss": 0.5044,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4362
    },
    {
      "epoch": 0.5932825673103073,
      "grad_norm": 0.373046875,
      "learning_rate": 8.493594850420537e-06,
      "loss": 0.5015,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4363
    },
    {
      "epoch": 0.593418547729127,
      "grad_norm": 0.30859375,
      "learning_rate": 8.488901439136681e-06,
      "loss": 0.5877,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4364
    },
    {
      "epoch": 0.5935545281479467,
      "grad_norm": 0.66796875,
      "learning_rate": 8.484208368472916e-06,
      "loss": 0.6864,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4365
    },
    {
      "epoch": 0.5936905085667664,
      "grad_norm": 0.37109375,
      "learning_rate": 8.479515639487115e-06,
      "loss": 0.5514,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4366
    },
    {
      "epoch": 0.5938264889855861,
      "grad_norm": 0.6875,
      "learning_rate": 8.474823253237081e-06,
      "loss": 0.6877,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4367
    },
    {
      "epoch": 0.5939624694044058,
      "grad_norm": 0.41796875,
      "learning_rate": 8.470131210780532e-06,
      "loss": 0.8192,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4368
    },
    {
      "epoch": 0.5940984498232255,
      "grad_norm": 0.2041015625,
      "learning_rate": 8.465439513175111e-06,
      "loss": 0.3698,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4369
    },
    {
      "epoch": 0.5942344302420451,
      "grad_norm": 0.396484375,
      "learning_rate": 8.460748161478386e-06,
      "loss": 0.625,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4370
    },
    {
      "epoch": 0.5943704106608648,
      "grad_norm": 0.48828125,
      "learning_rate": 8.456057156747847e-06,
      "loss": 0.6188,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4371
    },
    {
      "epoch": 0.5945063910796845,
      "grad_norm": 0.45703125,
      "learning_rate": 8.451366500040898e-06,
      "loss": 0.6388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4372
    },
    {
      "epoch": 0.5946423714985042,
      "grad_norm": 0.7578125,
      "learning_rate": 8.446676192414875e-06,
      "loss": 0.6593,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4373
    },
    {
      "epoch": 0.5947783519173239,
      "grad_norm": 0.33203125,
      "learning_rate": 8.44198623492703e-06,
      "loss": 0.6131,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4374
    },
    {
      "epoch": 0.5949143323361435,
      "grad_norm": 0.263671875,
      "learning_rate": 8.437296628634535e-06,
      "loss": 0.4801,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4375
    },
    {
      "epoch": 0.5950503127549633,
      "grad_norm": 0.33203125,
      "learning_rate": 8.432607374594484e-06,
      "loss": 0.569,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4376
    },
    {
      "epoch": 0.595186293173783,
      "grad_norm": 0.46484375,
      "learning_rate": 8.427918473863894e-06,
      "loss": 0.6616,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4377
    },
    {
      "epoch": 0.5953222735926027,
      "grad_norm": 0.337890625,
      "learning_rate": 8.423229927499704e-06,
      "loss": 0.6427,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4378
    },
    {
      "epoch": 0.5954582540114224,
      "grad_norm": 0.296875,
      "learning_rate": 8.418541736558763e-06,
      "loss": 0.6318,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4379
    },
    {
      "epoch": 0.5955942344302421,
      "grad_norm": 0.279296875,
      "learning_rate": 8.41385390209785e-06,
      "loss": 0.508,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4380
    },
    {
      "epoch": 0.5957302148490617,
      "grad_norm": 0.55859375,
      "learning_rate": 8.40916642517366e-06,
      "loss": 0.6634,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4381
    },
    {
      "epoch": 0.5958661952678814,
      "grad_norm": 0.498046875,
      "learning_rate": 8.40447930684281e-06,
      "loss": 0.5083,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4382
    },
    {
      "epoch": 0.5960021756867011,
      "grad_norm": 0.40234375,
      "learning_rate": 8.399792548161827e-06,
      "loss": 0.6497,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4383
    },
    {
      "epoch": 0.5961381561055208,
      "grad_norm": 0.255859375,
      "learning_rate": 8.395106150187171e-06,
      "loss": 0.4961,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4384
    },
    {
      "epoch": 0.5962741365243405,
      "grad_norm": 0.5078125,
      "learning_rate": 8.39042011397521e-06,
      "loss": 0.9159,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4385
    },
    {
      "epoch": 0.5964101169431602,
      "grad_norm": 0.412109375,
      "learning_rate": 8.385734440582232e-06,
      "loss": 0.7658,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4386
    },
    {
      "epoch": 0.5965460973619798,
      "grad_norm": 0.2431640625,
      "learning_rate": 8.38104913106445e-06,
      "loss": 0.4033,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4387
    },
    {
      "epoch": 0.5966820777807995,
      "grad_norm": 0.30859375,
      "learning_rate": 8.376364186477986e-06,
      "loss": 0.6517,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4388
    },
    {
      "epoch": 0.5968180581996193,
      "grad_norm": 0.41015625,
      "learning_rate": 8.371679607878884e-06,
      "loss": 0.5265,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4389
    },
    {
      "epoch": 0.596954038618439,
      "grad_norm": 0.35546875,
      "learning_rate": 8.366995396323107e-06,
      "loss": 0.5065,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4390
    },
    {
      "epoch": 0.5970900190372587,
      "grad_norm": 0.314453125,
      "learning_rate": 8.362311552866534e-06,
      "loss": 0.6048,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4391
    },
    {
      "epoch": 0.5972259994560783,
      "grad_norm": 0.37890625,
      "learning_rate": 8.357628078564958e-06,
      "loss": 0.5752,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4392
    },
    {
      "epoch": 0.597361979874898,
      "grad_norm": 0.259765625,
      "learning_rate": 8.352944974474094e-06,
      "loss": 0.4359,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4393
    },
    {
      "epoch": 0.5974979602937177,
      "grad_norm": 0.458984375,
      "learning_rate": 8.348262241649569e-06,
      "loss": 0.5798,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4394
    },
    {
      "epoch": 0.5976339407125374,
      "grad_norm": 0.64453125,
      "learning_rate": 8.34357988114693e-06,
      "loss": 0.6753,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4395
    },
    {
      "epoch": 0.5977699211313571,
      "grad_norm": 0.427734375,
      "learning_rate": 8.338897894021637e-06,
      "loss": 0.6468,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4396
    },
    {
      "epoch": 0.5979059015501768,
      "grad_norm": 0.62109375,
      "learning_rate": 8.334216281329067e-06,
      "loss": 0.5547,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4397
    },
    {
      "epoch": 0.5980418819689964,
      "grad_norm": 0.359375,
      "learning_rate": 8.329535044124514e-06,
      "loss": 0.7944,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4398
    },
    {
      "epoch": 0.5981778623878161,
      "grad_norm": 0.404296875,
      "learning_rate": 8.324854183463184e-06,
      "loss": 0.7181,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4399
    },
    {
      "epoch": 0.5983138428066358,
      "grad_norm": 0.357421875,
      "learning_rate": 8.320173700400202e-06,
      "loss": 0.4834,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4400
    },
    {
      "epoch": 0.5984498232254555,
      "grad_norm": 0.5234375,
      "learning_rate": 8.315493595990608e-06,
      "loss": 0.5768,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4401
    },
    {
      "epoch": 0.5985858036442753,
      "grad_norm": 0.234375,
      "learning_rate": 8.310813871289349e-06,
      "loss": 0.4102,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4402
    },
    {
      "epoch": 0.598721784063095,
      "grad_norm": 0.412109375,
      "learning_rate": 8.306134527351296e-06,
      "loss": 0.6398,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4403
    },
    {
      "epoch": 0.5988577644819146,
      "grad_norm": 0.296875,
      "learning_rate": 8.30145556523123e-06,
      "loss": 0.3846,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4404
    },
    {
      "epoch": 0.5989937449007343,
      "grad_norm": 1.515625,
      "learning_rate": 8.296776985983848e-06,
      "loss": 0.791,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4405
    },
    {
      "epoch": 0.599129725319554,
      "grad_norm": 0.44921875,
      "learning_rate": 8.292098790663756e-06,
      "loss": 0.5479,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4406
    },
    {
      "epoch": 0.5992657057383737,
      "grad_norm": 0.32421875,
      "learning_rate": 8.287420980325477e-06,
      "loss": 0.6147,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4407
    },
    {
      "epoch": 0.5994016861571934,
      "grad_norm": 0.3203125,
      "learning_rate": 8.282743556023447e-06,
      "loss": 0.5509,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4408
    },
    {
      "epoch": 0.599537666576013,
      "grad_norm": 0.365234375,
      "learning_rate": 8.278066518812015e-06,
      "loss": 0.6807,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4409
    },
    {
      "epoch": 0.5996736469948327,
      "grad_norm": 0.30078125,
      "learning_rate": 8.273389869745442e-06,
      "loss": 0.5601,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4410
    },
    {
      "epoch": 0.5998096274136524,
      "grad_norm": 0.33203125,
      "learning_rate": 8.268713609877905e-06,
      "loss": 0.603,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4411
    },
    {
      "epoch": 0.5999456078324721,
      "grad_norm": 0.37890625,
      "learning_rate": 8.264037740263484e-06,
      "loss": 0.6707,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4412
    },
    {
      "epoch": 0.6000815882512918,
      "grad_norm": 0.2890625,
      "learning_rate": 8.259362261956181e-06,
      "loss": 0.3776,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4413
    },
    {
      "epoch": 0.6002175686701116,
      "grad_norm": 0.37109375,
      "learning_rate": 8.254687176009906e-06,
      "loss": 0.4992,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4414
    },
    {
      "epoch": 0.6003535490889312,
      "grad_norm": 0.59765625,
      "learning_rate": 8.250012483478478e-06,
      "loss": 0.5479,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4415
    },
    {
      "epoch": 0.6004895295077509,
      "grad_norm": 0.34765625,
      "learning_rate": 8.245338185415634e-06,
      "loss": 0.6024,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4416
    },
    {
      "epoch": 0.6006255099265706,
      "grad_norm": 0.625,
      "learning_rate": 8.240664282875013e-06,
      "loss": 0.8981,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4417
    },
    {
      "epoch": 0.6007614903453903,
      "grad_norm": 0.58984375,
      "learning_rate": 8.235990776910174e-06,
      "loss": 0.5978,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4418
    },
    {
      "epoch": 0.60089747076421,
      "grad_norm": 0.3046875,
      "learning_rate": 8.231317668574581e-06,
      "loss": 0.6204,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4419
    },
    {
      "epoch": 0.6010334511830296,
      "grad_norm": 0.443359375,
      "learning_rate": 8.22664495892161e-06,
      "loss": 0.7791,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4420
    },
    {
      "epoch": 0.6011694316018493,
      "grad_norm": 0.416015625,
      "learning_rate": 8.221972649004546e-06,
      "loss": 0.6108,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4421
    },
    {
      "epoch": 0.601305412020669,
      "grad_norm": 0.326171875,
      "learning_rate": 8.217300739876585e-06,
      "loss": 0.6909,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4422
    },
    {
      "epoch": 0.6014413924394887,
      "grad_norm": 0.251953125,
      "learning_rate": 8.212629232590832e-06,
      "loss": 0.5524,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4423
    },
    {
      "epoch": 0.6015773728583084,
      "grad_norm": 0.75390625,
      "learning_rate": 8.207958128200304e-06,
      "loss": 0.8994,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4424
    },
    {
      "epoch": 0.601713353277128,
      "grad_norm": 0.396484375,
      "learning_rate": 8.203287427757923e-06,
      "loss": 0.7098,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4425
    },
    {
      "epoch": 0.6018493336959477,
      "grad_norm": 0.61328125,
      "learning_rate": 8.198617132316522e-06,
      "loss": 0.9502,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4426
    },
    {
      "epoch": 0.6019853141147675,
      "grad_norm": 0.3828125,
      "learning_rate": 8.193947242928844e-06,
      "loss": 0.8159,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4427
    },
    {
      "epoch": 0.6021212945335872,
      "grad_norm": 0.212890625,
      "learning_rate": 8.189277760647537e-06,
      "loss": 0.3991,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4428
    },
    {
      "epoch": 0.6022572749524069,
      "grad_norm": 0.255859375,
      "learning_rate": 8.184608686525162e-06,
      "loss": 0.6178,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4429
    },
    {
      "epoch": 0.6023932553712266,
      "grad_norm": 0.412109375,
      "learning_rate": 8.179940021614184e-06,
      "loss": 0.6816,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4430
    },
    {
      "epoch": 0.6025292357900462,
      "grad_norm": 0.232421875,
      "learning_rate": 8.175271766966977e-06,
      "loss": 0.4289,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4431
    },
    {
      "epoch": 0.6026652162088659,
      "grad_norm": 0.3515625,
      "learning_rate": 8.170603923635823e-06,
      "loss": 0.7526,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4432
    },
    {
      "epoch": 0.6028011966276856,
      "grad_norm": 0.51171875,
      "learning_rate": 8.165936492672912e-06,
      "loss": 0.5472,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4433
    },
    {
      "epoch": 0.6029371770465053,
      "grad_norm": 0.439453125,
      "learning_rate": 8.161269475130338e-06,
      "loss": 0.8791,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4434
    },
    {
      "epoch": 0.603073157465325,
      "grad_norm": 0.212890625,
      "learning_rate": 8.156602872060104e-06,
      "loss": 0.306,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4435
    },
    {
      "epoch": 0.6032091378841447,
      "grad_norm": 0.6015625,
      "learning_rate": 8.151936684514119e-06,
      "loss": 0.5602,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4436
    },
    {
      "epoch": 0.6033451183029643,
      "grad_norm": 0.287109375,
      "learning_rate": 8.147270913544202e-06,
      "loss": 0.5411,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4437
    },
    {
      "epoch": 0.603481098721784,
      "grad_norm": 0.251953125,
      "learning_rate": 8.142605560202073e-06,
      "loss": 0.5802,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4438
    },
    {
      "epoch": 0.6036170791406038,
      "grad_norm": 0.3046875,
      "learning_rate": 8.137940625539357e-06,
      "loss": 0.5493,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4439
    },
    {
      "epoch": 0.6037530595594235,
      "grad_norm": 0.443359375,
      "learning_rate": 8.133276110607593e-06,
      "loss": 0.6914,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4440
    },
    {
      "epoch": 0.6038890399782432,
      "grad_norm": 0.4453125,
      "learning_rate": 8.128612016458214e-06,
      "loss": 0.7244,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4441
    },
    {
      "epoch": 0.6040250203970629,
      "grad_norm": 0.341796875,
      "learning_rate": 8.123948344142568e-06,
      "loss": 0.715,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4442
    },
    {
      "epoch": 0.6041610008158825,
      "grad_norm": 0.50390625,
      "learning_rate": 8.119285094711901e-06,
      "loss": 0.5348,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4443
    },
    {
      "epoch": 0.6042969812347022,
      "grad_norm": 0.240234375,
      "learning_rate": 8.114622269217368e-06,
      "loss": 0.4035,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4444
    },
    {
      "epoch": 0.6044329616535219,
      "grad_norm": 0.3828125,
      "learning_rate": 8.109959868710026e-06,
      "loss": 0.6265,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4445
    },
    {
      "epoch": 0.6045689420723416,
      "grad_norm": 0.37890625,
      "learning_rate": 8.105297894240838e-06,
      "loss": 0.6971,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4446
    },
    {
      "epoch": 0.6047049224911613,
      "grad_norm": 0.328125,
      "learning_rate": 8.100636346860668e-06,
      "loss": 0.5187,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4447
    },
    {
      "epoch": 0.6048409029099809,
      "grad_norm": 0.373046875,
      "learning_rate": 8.095975227620289e-06,
      "loss": 0.6637,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4448
    },
    {
      "epoch": 0.6049768833288006,
      "grad_norm": 0.3125,
      "learning_rate": 8.091314537570371e-06,
      "loss": 0.561,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4449
    },
    {
      "epoch": 0.6051128637476203,
      "grad_norm": 0.546875,
      "learning_rate": 8.086654277761493e-06,
      "loss": 0.6434,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4450
    },
    {
      "epoch": 0.60524884416644,
      "grad_norm": 0.5234375,
      "learning_rate": 8.081994449244132e-06,
      "loss": 0.9176,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4451
    },
    {
      "epoch": 0.6053848245852598,
      "grad_norm": 0.57421875,
      "learning_rate": 8.077335053068673e-06,
      "loss": 0.724,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4452
    },
    {
      "epoch": 0.6055208050040795,
      "grad_norm": 0.357421875,
      "learning_rate": 8.0726760902854e-06,
      "loss": 0.5744,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4453
    },
    {
      "epoch": 0.6056567854228991,
      "grad_norm": 2.796875,
      "learning_rate": 8.0680175619445e-06,
      "loss": 0.8841,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4454
    },
    {
      "epoch": 0.6057927658417188,
      "grad_norm": 1.296875,
      "learning_rate": 8.063359469096061e-06,
      "loss": 0.7038,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4455
    },
    {
      "epoch": 0.6059287462605385,
      "grad_norm": 0.322265625,
      "learning_rate": 8.058701812790076e-06,
      "loss": 0.638,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4456
    },
    {
      "epoch": 0.6060647266793582,
      "grad_norm": 0.5,
      "learning_rate": 8.054044594076435e-06,
      "loss": 0.6185,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4457
    },
    {
      "epoch": 0.6062007070981779,
      "grad_norm": 0.328125,
      "learning_rate": 8.049387814004936e-06,
      "loss": 0.5716,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4458
    },
    {
      "epoch": 0.6063366875169975,
      "grad_norm": 0.3125,
      "learning_rate": 8.044731473625272e-06,
      "loss": 0.6408,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4459
    },
    {
      "epoch": 0.6064726679358172,
      "grad_norm": 0.515625,
      "learning_rate": 8.040075573987038e-06,
      "loss": 0.638,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4460
    },
    {
      "epoch": 0.6066086483546369,
      "grad_norm": 0.57421875,
      "learning_rate": 8.035420116139732e-06,
      "loss": 0.7892,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4461
    },
    {
      "epoch": 0.6067446287734566,
      "grad_norm": 0.37890625,
      "learning_rate": 8.030765101132751e-06,
      "loss": 0.6732,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4462
    },
    {
      "epoch": 0.6068806091922763,
      "grad_norm": 0.373046875,
      "learning_rate": 8.026110530015391e-06,
      "loss": 0.7114,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4463
    },
    {
      "epoch": 0.607016589611096,
      "grad_norm": 0.57421875,
      "learning_rate": 8.021456403836853e-06,
      "loss": 0.4694,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4464
    },
    {
      "epoch": 0.6071525700299157,
      "grad_norm": 0.4921875,
      "learning_rate": 8.01680272364623e-06,
      "loss": 0.7061,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4465
    },
    {
      "epoch": 0.6072885504487354,
      "grad_norm": 0.30078125,
      "learning_rate": 8.01214949049252e-06,
      "loss": 0.6169,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4466
    },
    {
      "epoch": 0.6074245308675551,
      "grad_norm": 0.6640625,
      "learning_rate": 8.00749670542462e-06,
      "loss": 0.6969,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4467
    },
    {
      "epoch": 0.6075605112863748,
      "grad_norm": 0.4140625,
      "learning_rate": 8.00284436949132e-06,
      "loss": 0.5095,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4468
    },
    {
      "epoch": 0.6076964917051945,
      "grad_norm": 0.326171875,
      "learning_rate": 7.99819248374132e-06,
      "loss": 0.6404,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4469
    },
    {
      "epoch": 0.6078324721240141,
      "grad_norm": 0.6015625,
      "learning_rate": 7.993541049223205e-06,
      "loss": 0.8314,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4470
    },
    {
      "epoch": 0.6079684525428338,
      "grad_norm": 0.4609375,
      "learning_rate": 7.988890066985471e-06,
      "loss": 0.599,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4471
    },
    {
      "epoch": 0.6081044329616535,
      "grad_norm": 0.25390625,
      "learning_rate": 7.984239538076503e-06,
      "loss": 0.4587,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4472
    },
    {
      "epoch": 0.6082404133804732,
      "grad_norm": 0.44921875,
      "learning_rate": 7.97958946354459e-06,
      "loss": 0.5008,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4473
    },
    {
      "epoch": 0.6083763937992929,
      "grad_norm": 0.396484375,
      "learning_rate": 7.974939844437912e-06,
      "loss": 0.529,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4474
    },
    {
      "epoch": 0.6085123742181126,
      "grad_norm": 0.31640625,
      "learning_rate": 7.97029068180455e-06,
      "loss": 0.6169,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4475
    },
    {
      "epoch": 0.6086483546369322,
      "grad_norm": 0.32421875,
      "learning_rate": 7.965641976692486e-06,
      "loss": 0.5291,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4476
    },
    {
      "epoch": 0.608784335055752,
      "grad_norm": 0.65625,
      "learning_rate": 7.960993730149592e-06,
      "loss": 0.6848,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4477
    },
    {
      "epoch": 0.6089203154745717,
      "grad_norm": 0.58984375,
      "learning_rate": 7.956345943223639e-06,
      "loss": 0.7736,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4478
    },
    {
      "epoch": 0.6090562958933914,
      "grad_norm": 0.455078125,
      "learning_rate": 7.951698616962299e-06,
      "loss": 0.5688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4479
    },
    {
      "epoch": 0.6091922763122111,
      "grad_norm": 0.39453125,
      "learning_rate": 7.947051752413131e-06,
      "loss": 0.7234,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4480
    },
    {
      "epoch": 0.6093282567310307,
      "grad_norm": 0.486328125,
      "learning_rate": 7.942405350623597e-06,
      "loss": 0.6217,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4481
    },
    {
      "epoch": 0.6094642371498504,
      "grad_norm": 0.6015625,
      "learning_rate": 7.937759412641055e-06,
      "loss": 0.6452,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4482
    },
    {
      "epoch": 0.6096002175686701,
      "grad_norm": 0.3359375,
      "learning_rate": 7.933113939512754e-06,
      "loss": 0.6307,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4483
    },
    {
      "epoch": 0.6097361979874898,
      "grad_norm": 0.388671875,
      "learning_rate": 7.92846893228584e-06,
      "loss": 0.5202,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4484
    },
    {
      "epoch": 0.6098721784063095,
      "grad_norm": 0.384765625,
      "learning_rate": 7.923824392007357e-06,
      "loss": 0.7417,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4485
    },
    {
      "epoch": 0.6100081588251292,
      "grad_norm": 0.4140625,
      "learning_rate": 7.91918031972424e-06,
      "loss": 0.555,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4486
    },
    {
      "epoch": 0.6101441392439488,
      "grad_norm": 0.423828125,
      "learning_rate": 7.914536716483317e-06,
      "loss": 0.7526,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4487
    },
    {
      "epoch": 0.6102801196627685,
      "grad_norm": 0.4296875,
      "learning_rate": 7.909893583331316e-06,
      "loss": 0.6947,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4488
    },
    {
      "epoch": 0.6104161000815882,
      "grad_norm": 0.41796875,
      "learning_rate": 7.905250921314857e-06,
      "loss": 0.6252,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4489
    },
    {
      "epoch": 0.610552080500408,
      "grad_norm": 0.62109375,
      "learning_rate": 7.900608731480449e-06,
      "loss": 0.5804,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4490
    },
    {
      "epoch": 0.6106880609192277,
      "grad_norm": 0.2138671875,
      "learning_rate": 7.8959670148745e-06,
      "loss": 0.4299,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4491
    },
    {
      "epoch": 0.6108240413380474,
      "grad_norm": 0.50390625,
      "learning_rate": 7.89132577254331e-06,
      "loss": 0.6951,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4492
    },
    {
      "epoch": 0.610960021756867,
      "grad_norm": 0.32421875,
      "learning_rate": 7.886685005533073e-06,
      "loss": 0.7323,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4493
    },
    {
      "epoch": 0.6110960021756867,
      "grad_norm": 0.328125,
      "learning_rate": 7.882044714889872e-06,
      "loss": 0.4743,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4494
    },
    {
      "epoch": 0.6112319825945064,
      "grad_norm": 0.376953125,
      "learning_rate": 7.877404901659687e-06,
      "loss": 0.5672,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4495
    },
    {
      "epoch": 0.6113679630133261,
      "grad_norm": 0.349609375,
      "learning_rate": 7.872765566888387e-06,
      "loss": 0.6159,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4496
    },
    {
      "epoch": 0.6115039434321458,
      "grad_norm": 0.51171875,
      "learning_rate": 7.868126711621737e-06,
      "loss": 0.6764,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4497
    },
    {
      "epoch": 0.6116399238509654,
      "grad_norm": 0.46875,
      "learning_rate": 7.86348833690539e-06,
      "loss": 0.6473,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4498
    },
    {
      "epoch": 0.6117759042697851,
      "grad_norm": 0.609375,
      "learning_rate": 7.858850443784893e-06,
      "loss": 0.6618,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4499
    },
    {
      "epoch": 0.6119118846886048,
      "grad_norm": 0.28125,
      "learning_rate": 7.854213033305685e-06,
      "loss": 0.502,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4500
    },
    {
      "epoch": 0.6120478651074245,
      "grad_norm": 0.86328125,
      "learning_rate": 7.849576106513095e-06,
      "loss": 0.7354,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4501
    },
    {
      "epoch": 0.6121838455262443,
      "grad_norm": 0.69140625,
      "learning_rate": 7.844939664452341e-06,
      "loss": 0.6014,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4502
    },
    {
      "epoch": 0.612319825945064,
      "grad_norm": 0.361328125,
      "learning_rate": 7.840303708168533e-06,
      "loss": 0.5962,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4503
    },
    {
      "epoch": 0.6124558063638836,
      "grad_norm": 0.26953125,
      "learning_rate": 7.835668238706674e-06,
      "loss": 0.5628,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4504
    },
    {
      "epoch": 0.6125917867827033,
      "grad_norm": 0.400390625,
      "learning_rate": 7.831033257111656e-06,
      "loss": 0.4979,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4505
    },
    {
      "epoch": 0.612727767201523,
      "grad_norm": 0.44140625,
      "learning_rate": 7.826398764428261e-06,
      "loss": 0.8197,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4506
    },
    {
      "epoch": 0.6128637476203427,
      "grad_norm": 0.73828125,
      "learning_rate": 7.821764761701158e-06,
      "loss": 0.6486,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4507
    },
    {
      "epoch": 0.6129997280391624,
      "grad_norm": 0.32421875,
      "learning_rate": 7.817131249974908e-06,
      "loss": 0.6346,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4508
    },
    {
      "epoch": 0.613135708457982,
      "grad_norm": 0.53515625,
      "learning_rate": 7.812498230293964e-06,
      "loss": 0.7145,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4509
    },
    {
      "epoch": 0.6132716888768017,
      "grad_norm": 0.361328125,
      "learning_rate": 7.807865703702661e-06,
      "loss": 0.6916,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4510
    },
    {
      "epoch": 0.6134076692956214,
      "grad_norm": 0.34375,
      "learning_rate": 7.803233671245232e-06,
      "loss": 0.4925,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4511
    },
    {
      "epoch": 0.6135436497144411,
      "grad_norm": 0.5703125,
      "learning_rate": 7.79860213396579e-06,
      "loss": 0.6647,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4512
    },
    {
      "epoch": 0.6136796301332608,
      "grad_norm": 0.462890625,
      "learning_rate": 7.793971092908343e-06,
      "loss": 0.7985,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4513
    },
    {
      "epoch": 0.6138156105520804,
      "grad_norm": 0.5,
      "learning_rate": 7.789340549116782e-06,
      "loss": 0.7329,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4514
    },
    {
      "epoch": 0.6139515909709002,
      "grad_norm": 0.57421875,
      "learning_rate": 7.784710503634888e-06,
      "loss": 0.6904,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4515
    },
    {
      "epoch": 0.6140875713897199,
      "grad_norm": 0.31640625,
      "learning_rate": 7.78008095750633e-06,
      "loss": 0.6211,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4516
    },
    {
      "epoch": 0.6142235518085396,
      "grad_norm": 0.408203125,
      "learning_rate": 7.775451911774667e-06,
      "loss": 0.6616,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4517
    },
    {
      "epoch": 0.6143595322273593,
      "grad_norm": 0.482421875,
      "learning_rate": 7.770823367483342e-06,
      "loss": 0.7207,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4518
    },
    {
      "epoch": 0.614495512646179,
      "grad_norm": 0.390625,
      "learning_rate": 7.766195325675681e-06,
      "loss": 0.661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4519
    },
    {
      "epoch": 0.6146314930649986,
      "grad_norm": 0.5078125,
      "learning_rate": 7.761567787394906e-06,
      "loss": 0.7305,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4520
    },
    {
      "epoch": 0.6147674734838183,
      "grad_norm": 0.53515625,
      "learning_rate": 7.756940753684117e-06,
      "loss": 0.7301,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4521
    },
    {
      "epoch": 0.614903453902638,
      "grad_norm": 0.68359375,
      "learning_rate": 7.752314225586307e-06,
      "loss": 0.5573,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4522
    },
    {
      "epoch": 0.6150394343214577,
      "grad_norm": 0.337890625,
      "learning_rate": 7.747688204144352e-06,
      "loss": 0.4884,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4523
    },
    {
      "epoch": 0.6151754147402774,
      "grad_norm": 0.404296875,
      "learning_rate": 7.743062690401014e-06,
      "loss": 0.5399,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4524
    },
    {
      "epoch": 0.615311395159097,
      "grad_norm": 0.314453125,
      "learning_rate": 7.738437685398937e-06,
      "loss": 0.5446,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4525
    },
    {
      "epoch": 0.6154473755779167,
      "grad_norm": 0.3203125,
      "learning_rate": 7.733813190180655e-06,
      "loss": 0.5794,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4526
    },
    {
      "epoch": 0.6155833559967365,
      "grad_norm": 0.3515625,
      "learning_rate": 7.729189205788588e-06,
      "loss": 0.5954,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4527
    },
    {
      "epoch": 0.6157193364155562,
      "grad_norm": 0.435546875,
      "learning_rate": 7.724565733265036e-06,
      "loss": 0.8335,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4528
    },
    {
      "epoch": 0.6158553168343759,
      "grad_norm": 0.56640625,
      "learning_rate": 7.719942773652188e-06,
      "loss": 0.5558,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4529
    },
    {
      "epoch": 0.6159912972531956,
      "grad_norm": 0.40234375,
      "learning_rate": 7.715320327992115e-06,
      "loss": 0.888,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4530
    },
    {
      "epoch": 0.6161272776720152,
      "grad_norm": 0.2890625,
      "learning_rate": 7.710698397326772e-06,
      "loss": 0.3887,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4531
    },
    {
      "epoch": 0.6162632580908349,
      "grad_norm": 0.42578125,
      "learning_rate": 7.706076982698e-06,
      "loss": 0.6571,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4532
    },
    {
      "epoch": 0.6163992385096546,
      "grad_norm": 0.279296875,
      "learning_rate": 7.701456085147522e-06,
      "loss": 0.4595,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4533
    },
    {
      "epoch": 0.6165352189284743,
      "grad_norm": 0.4609375,
      "learning_rate": 7.696835705716945e-06,
      "loss": 0.8136,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4534
    },
    {
      "epoch": 0.616671199347294,
      "grad_norm": 0.330078125,
      "learning_rate": 7.692215845447757e-06,
      "loss": 0.515,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4535
    },
    {
      "epoch": 0.6168071797661137,
      "grad_norm": 0.470703125,
      "learning_rate": 7.687596505381334e-06,
      "loss": 0.484,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4536
    },
    {
      "epoch": 0.6169431601849333,
      "grad_norm": 0.5859375,
      "learning_rate": 7.68297768655893e-06,
      "loss": 0.687,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4537
    },
    {
      "epoch": 0.617079140603753,
      "grad_norm": 0.3125,
      "learning_rate": 7.678359390021683e-06,
      "loss": 0.5786,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4538
    },
    {
      "epoch": 0.6172151210225727,
      "grad_norm": 0.392578125,
      "learning_rate": 7.673741616810615e-06,
      "loss": 0.73,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4539
    },
    {
      "epoch": 0.6173511014413925,
      "grad_norm": 0.3828125,
      "learning_rate": 7.669124367966626e-06,
      "loss": 0.8057,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4540
    },
    {
      "epoch": 0.6174870818602122,
      "grad_norm": 0.65625,
      "learning_rate": 7.664507644530504e-06,
      "loss": 0.3857,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4541
    },
    {
      "epoch": 0.6176230622790319,
      "grad_norm": 0.408203125,
      "learning_rate": 7.659891447542912e-06,
      "loss": 0.6102,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4542
    },
    {
      "epoch": 0.6177590426978515,
      "grad_norm": 0.3125,
      "learning_rate": 7.655275778044398e-06,
      "loss": 0.5671,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4543
    },
    {
      "epoch": 0.6178950231166712,
      "grad_norm": 0.29296875,
      "learning_rate": 7.65066063707539e-06,
      "loss": 0.5622,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4544
    },
    {
      "epoch": 0.6180310035354909,
      "grad_norm": 0.58203125,
      "learning_rate": 7.646046025676198e-06,
      "loss": 0.7386,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4545
    },
    {
      "epoch": 0.6181669839543106,
      "grad_norm": 0.8359375,
      "learning_rate": 7.64143194488701e-06,
      "loss": 0.6685,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4546
    },
    {
      "epoch": 0.6183029643731303,
      "grad_norm": 0.28125,
      "learning_rate": 7.6368183957479e-06,
      "loss": 0.5795,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4547
    },
    {
      "epoch": 0.6184389447919499,
      "grad_norm": 0.375,
      "learning_rate": 7.632205379298816e-06,
      "loss": 0.6388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4548
    },
    {
      "epoch": 0.6185749252107696,
      "grad_norm": 0.43359375,
      "learning_rate": 7.6275928965795875e-06,
      "loss": 0.7165,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4549
    },
    {
      "epoch": 0.6187109056295893,
      "grad_norm": 0.294921875,
      "learning_rate": 7.622980948629925e-06,
      "loss": 0.488,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4550
    },
    {
      "epoch": 0.618846886048409,
      "grad_norm": 0.86328125,
      "learning_rate": 7.618369536489418e-06,
      "loss": 0.6094,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4551
    },
    {
      "epoch": 0.6189828664672287,
      "grad_norm": 0.35546875,
      "learning_rate": 7.613758661197537e-06,
      "loss": 0.6445,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4552
    },
    {
      "epoch": 0.6191188468860485,
      "grad_norm": 0.396484375,
      "learning_rate": 7.609148323793627e-06,
      "loss": 0.8978,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4553
    },
    {
      "epoch": 0.6192548273048681,
      "grad_norm": 0.349609375,
      "learning_rate": 7.604538525316915e-06,
      "loss": 0.6123,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4554
    },
    {
      "epoch": 0.6193908077236878,
      "grad_norm": 0.283203125,
      "learning_rate": 7.599929266806508e-06,
      "loss": 0.5597,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4555
    },
    {
      "epoch": 0.6195267881425075,
      "grad_norm": 0.38671875,
      "learning_rate": 7.595320549301385e-06,
      "loss": 0.6092,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4556
    },
    {
      "epoch": 0.6196627685613272,
      "grad_norm": 0.4921875,
      "learning_rate": 7.5907123738404124e-06,
      "loss": 0.8594,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4557
    },
    {
      "epoch": 0.6197987489801469,
      "grad_norm": 0.25390625,
      "learning_rate": 7.586104741462326e-06,
      "loss": 0.4398,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4558
    },
    {
      "epoch": 0.6199347293989665,
      "grad_norm": 0.5546875,
      "learning_rate": 7.5814976532057435e-06,
      "loss": 0.8172,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4559
    },
    {
      "epoch": 0.6200707098177862,
      "grad_norm": 0.40625,
      "learning_rate": 7.576891110109159e-06,
      "loss": 0.5931,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4560
    },
    {
      "epoch": 0.6202066902366059,
      "grad_norm": 0.322265625,
      "learning_rate": 7.572285113210944e-06,
      "loss": 0.5111,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4561
    },
    {
      "epoch": 0.6203426706554256,
      "grad_norm": 0.275390625,
      "learning_rate": 7.567679663549346e-06,
      "loss": 0.6278,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4562
    },
    {
      "epoch": 0.6204786510742453,
      "grad_norm": 0.341796875,
      "learning_rate": 7.563074762162488e-06,
      "loss": 0.6668,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4563
    },
    {
      "epoch": 0.620614631493065,
      "grad_norm": 0.427734375,
      "learning_rate": 7.558470410088374e-06,
      "loss": 0.7054,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4564
    },
    {
      "epoch": 0.6207506119118847,
      "grad_norm": 0.294921875,
      "learning_rate": 7.5538666083648796e-06,
      "loss": 0.6108,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4565
    },
    {
      "epoch": 0.6208865923307044,
      "grad_norm": 0.357421875,
      "learning_rate": 7.549263358029758e-06,
      "loss": 0.6403,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4566
    },
    {
      "epoch": 0.6210225727495241,
      "grad_norm": 0.30078125,
      "learning_rate": 7.5446606601206395e-06,
      "loss": 0.5944,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4567
    },
    {
      "epoch": 0.6211585531683438,
      "grad_norm": 0.234375,
      "learning_rate": 7.540058515675026e-06,
      "loss": 0.3859,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4568
    },
    {
      "epoch": 0.6212945335871635,
      "grad_norm": 0.5859375,
      "learning_rate": 7.535456925730299e-06,
      "loss": 0.5885,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4569
    },
    {
      "epoch": 0.6214305140059831,
      "grad_norm": 0.25,
      "learning_rate": 7.530855891323713e-06,
      "loss": 0.425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4570
    },
    {
      "epoch": 0.6215664944248028,
      "grad_norm": 0.314453125,
      "learning_rate": 7.526255413492396e-06,
      "loss": 0.5581,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4571
    },
    {
      "epoch": 0.6217024748436225,
      "grad_norm": 0.51953125,
      "learning_rate": 7.521655493273354e-06,
      "loss": 0.7148,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4572
    },
    {
      "epoch": 0.6218384552624422,
      "grad_norm": 0.34375,
      "learning_rate": 7.517056131703464e-06,
      "loss": 0.5624,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4573
    },
    {
      "epoch": 0.6219744356812619,
      "grad_norm": 0.466796875,
      "learning_rate": 7.512457329819478e-06,
      "loss": 0.7578,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4574
    },
    {
      "epoch": 0.6221104161000816,
      "grad_norm": 0.494140625,
      "learning_rate": 7.507859088658024e-06,
      "loss": 0.8123,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4575
    },
    {
      "epoch": 0.6222463965189012,
      "grad_norm": 0.2353515625,
      "learning_rate": 7.5032614092556e-06,
      "loss": 0.4097,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4576
    },
    {
      "epoch": 0.6223823769377209,
      "grad_norm": 0.42578125,
      "learning_rate": 7.498664292648579e-06,
      "loss": 0.5464,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4577
    },
    {
      "epoch": 0.6225183573565407,
      "grad_norm": 0.71484375,
      "learning_rate": 7.494067739873209e-06,
      "loss": 0.8197,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4578
    },
    {
      "epoch": 0.6226543377753604,
      "grad_norm": 0.40625,
      "learning_rate": 7.489471751965607e-06,
      "loss": 0.6032,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4579
    },
    {
      "epoch": 0.6227903181941801,
      "grad_norm": 0.75,
      "learning_rate": 7.484876329961768e-06,
      "loss": 0.8659,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4580
    },
    {
      "epoch": 0.6229262986129998,
      "grad_norm": 0.3984375,
      "learning_rate": 7.480281474897553e-06,
      "loss": 0.6787,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4581
    },
    {
      "epoch": 0.6230622790318194,
      "grad_norm": 0.5390625,
      "learning_rate": 7.475687187808699e-06,
      "loss": 0.7,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4582
    },
    {
      "epoch": 0.6231982594506391,
      "grad_norm": 0.671875,
      "learning_rate": 7.471093469730817e-06,
      "loss": 0.619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4583
    },
    {
      "epoch": 0.6233342398694588,
      "grad_norm": 0.341796875,
      "learning_rate": 7.4665003216993835e-06,
      "loss": 0.6325,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4584
    },
    {
      "epoch": 0.6234702202882785,
      "grad_norm": 0.453125,
      "learning_rate": 7.461907744749754e-06,
      "loss": 0.6985,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4585
    },
    {
      "epoch": 0.6236062007070982,
      "grad_norm": 0.318359375,
      "learning_rate": 7.457315739917149e-06,
      "loss": 0.6453,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4586
    },
    {
      "epoch": 0.6237421811259178,
      "grad_norm": 0.70703125,
      "learning_rate": 7.452724308236663e-06,
      "loss": 0.6701,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4587
    },
    {
      "epoch": 0.6238781615447375,
      "grad_norm": 0.59765625,
      "learning_rate": 7.448133450743262e-06,
      "loss": 0.5715,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4588
    },
    {
      "epoch": 0.6240141419635572,
      "grad_norm": 0.296875,
      "learning_rate": 7.443543168471781e-06,
      "loss": 0.4954,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4589
    },
    {
      "epoch": 0.624150122382377,
      "grad_norm": 0.66015625,
      "learning_rate": 7.438953462456924e-06,
      "loss": 0.6969,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4590
    },
    {
      "epoch": 0.6242861028011967,
      "grad_norm": 0.400390625,
      "learning_rate": 7.434364333733268e-06,
      "loss": 0.812,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4591
    },
    {
      "epoch": 0.6244220832200164,
      "grad_norm": 0.40234375,
      "learning_rate": 7.42977578333526e-06,
      "loss": 0.675,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4592
    },
    {
      "epoch": 0.624558063638836,
      "grad_norm": 0.275390625,
      "learning_rate": 7.425187812297214e-06,
      "loss": 0.484,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4593
    },
    {
      "epoch": 0.6246940440576557,
      "grad_norm": 0.359375,
      "learning_rate": 7.420600421653314e-06,
      "loss": 0.6315,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4594
    },
    {
      "epoch": 0.6248300244764754,
      "grad_norm": 0.423828125,
      "learning_rate": 7.416013612437616e-06,
      "loss": 0.7021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4595
    },
    {
      "epoch": 0.6249660048952951,
      "grad_norm": 0.32421875,
      "learning_rate": 7.4114273856840405e-06,
      "loss": 0.661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4596
    },
    {
      "epoch": 0.6251019853141148,
      "grad_norm": 0.5546875,
      "learning_rate": 7.40684174242638e-06,
      "loss": 0.7837,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4597
    },
    {
      "epoch": 0.6252379657329344,
      "grad_norm": 0.55859375,
      "learning_rate": 7.4022566836982964e-06,
      "loss": 0.7965,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4598
    },
    {
      "epoch": 0.6253739461517541,
      "grad_norm": 0.43359375,
      "learning_rate": 7.3976722105333175e-06,
      "loss": 0.8158,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4599
    },
    {
      "epoch": 0.6255099265705738,
      "grad_norm": 0.392578125,
      "learning_rate": 7.393088323964836e-06,
      "loss": 0.772,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4600
    },
    {
      "epoch": 0.6256459069893935,
      "grad_norm": 0.361328125,
      "learning_rate": 7.38850502502612e-06,
      "loss": 0.6896,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4601
    },
    {
      "epoch": 0.6257818874082132,
      "grad_norm": 0.27734375,
      "learning_rate": 7.3839223147503e-06,
      "loss": 0.557,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4602
    },
    {
      "epoch": 0.625917867827033,
      "grad_norm": 0.484375,
      "learning_rate": 7.3793401941703745e-06,
      "loss": 0.6257,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4603
    },
    {
      "epoch": 0.6260538482458526,
      "grad_norm": 0.5546875,
      "learning_rate": 7.374758664319208e-06,
      "loss": 0.383,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4604
    },
    {
      "epoch": 0.6261898286646723,
      "grad_norm": 0.34765625,
      "learning_rate": 7.370177726229537e-06,
      "loss": 0.6084,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4605
    },
    {
      "epoch": 0.626325809083492,
      "grad_norm": 0.578125,
      "learning_rate": 7.3655973809339585e-06,
      "loss": 0.637,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4606
    },
    {
      "epoch": 0.6264617895023117,
      "grad_norm": 0.3671875,
      "learning_rate": 7.361017629464938e-06,
      "loss": 0.6559,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4607
    },
    {
      "epoch": 0.6265977699211314,
      "grad_norm": 0.42578125,
      "learning_rate": 7.356438472854809e-06,
      "loss": 0.661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4608
    },
    {
      "epoch": 0.626733750339951,
      "grad_norm": 0.38671875,
      "learning_rate": 7.3518599121357685e-06,
      "loss": 0.6393,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4609
    },
    {
      "epoch": 0.6268697307587707,
      "grad_norm": 0.400390625,
      "learning_rate": 7.34728194833988e-06,
      "loss": 0.7485,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4610
    },
    {
      "epoch": 0.6270057111775904,
      "grad_norm": 0.3203125,
      "learning_rate": 7.342704582499072e-06,
      "loss": 0.5846,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4611
    },
    {
      "epoch": 0.6271416915964101,
      "grad_norm": 0.33203125,
      "learning_rate": 7.33812781564514e-06,
      "loss": 0.6209,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4612
    },
    {
      "epoch": 0.6272776720152298,
      "grad_norm": 0.6015625,
      "learning_rate": 7.333551648809743e-06,
      "loss": 0.7131,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4613
    },
    {
      "epoch": 0.6274136524340495,
      "grad_norm": 0.3203125,
      "learning_rate": 7.328976083024404e-06,
      "loss": 0.576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4614
    },
    {
      "epoch": 0.6275496328528691,
      "grad_norm": 0.333984375,
      "learning_rate": 7.324401119320512e-06,
      "loss": 0.4836,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4615
    },
    {
      "epoch": 0.6276856132716889,
      "grad_norm": 0.388671875,
      "learning_rate": 7.319826758729318e-06,
      "loss": 0.6458,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4616
    },
    {
      "epoch": 0.6278215936905086,
      "grad_norm": 0.416015625,
      "learning_rate": 7.315253002281944e-06,
      "loss": 0.7266,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4617
    },
    {
      "epoch": 0.6279575741093283,
      "grad_norm": 0.8046875,
      "learning_rate": 7.310679851009366e-06,
      "loss": 0.8835,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4618
    },
    {
      "epoch": 0.628093554528148,
      "grad_norm": 0.333984375,
      "learning_rate": 7.306107305942428e-06,
      "loss": 0.5841,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4619
    },
    {
      "epoch": 0.6282295349469676,
      "grad_norm": 0.474609375,
      "learning_rate": 7.30153536811184e-06,
      "loss": 0.5556,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4620
    },
    {
      "epoch": 0.6283655153657873,
      "grad_norm": 0.38671875,
      "learning_rate": 7.2969640385481734e-06,
      "loss": 0.4364,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4621
    },
    {
      "epoch": 0.628501495784607,
      "grad_norm": 0.72265625,
      "learning_rate": 7.292393318281858e-06,
      "loss": 0.7785,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4622
    },
    {
      "epoch": 0.6286374762034267,
      "grad_norm": 0.63671875,
      "learning_rate": 7.287823208343192e-06,
      "loss": 0.4269,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4623
    },
    {
      "epoch": 0.6287734566222464,
      "grad_norm": 0.37109375,
      "learning_rate": 7.283253709762335e-06,
      "loss": 0.5229,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4624
    },
    {
      "epoch": 0.628909437041066,
      "grad_norm": 0.76171875,
      "learning_rate": 7.278684823569306e-06,
      "loss": 0.5793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4625
    },
    {
      "epoch": 0.6290454174598857,
      "grad_norm": 0.33203125,
      "learning_rate": 7.2741165507939905e-06,
      "loss": 0.5731,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4626
    },
    {
      "epoch": 0.6291813978787054,
      "grad_norm": 0.4375,
      "learning_rate": 7.26954889246613e-06,
      "loss": 0.8371,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4627
    },
    {
      "epoch": 0.6293173782975252,
      "grad_norm": 0.279296875,
      "learning_rate": 7.2649818496153335e-06,
      "loss": 0.5096,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4628
    },
    {
      "epoch": 0.6294533587163449,
      "grad_norm": 0.388671875,
      "learning_rate": 7.260415423271065e-06,
      "loss": 0.3589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4629
    },
    {
      "epoch": 0.6295893391351646,
      "grad_norm": 0.4609375,
      "learning_rate": 7.255849614462654e-06,
      "loss": 0.5508,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4630
    },
    {
      "epoch": 0.6297253195539843,
      "grad_norm": 0.48046875,
      "learning_rate": 7.2512844242192915e-06,
      "loss": 0.6992,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4631
    },
    {
      "epoch": 0.6298612999728039,
      "grad_norm": 0.4296875,
      "learning_rate": 7.2467198535700256e-06,
      "loss": 0.7077,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4632
    },
    {
      "epoch": 0.6299972803916236,
      "grad_norm": 0.55078125,
      "learning_rate": 7.242155903543766e-06,
      "loss": 0.6551,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4633
    },
    {
      "epoch": 0.6301332608104433,
      "grad_norm": 0.27734375,
      "learning_rate": 7.237592575169284e-06,
      "loss": 0.5555,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4634
    },
    {
      "epoch": 0.630269241229263,
      "grad_norm": 0.8046875,
      "learning_rate": 7.233029869475207e-06,
      "loss": 0.6818,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4635
    },
    {
      "epoch": 0.6304052216480827,
      "grad_norm": 0.3671875,
      "learning_rate": 7.2284677874900285e-06,
      "loss": 0.6087,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4636
    },
    {
      "epoch": 0.6305412020669023,
      "grad_norm": 0.66796875,
      "learning_rate": 7.223906330242097e-06,
      "loss": 0.5845,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4637
    },
    {
      "epoch": 0.630677182485722,
      "grad_norm": 0.392578125,
      "learning_rate": 7.219345498759616e-06,
      "loss": 0.8062,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4638
    },
    {
      "epoch": 0.6308131629045417,
      "grad_norm": 0.267578125,
      "learning_rate": 7.214785294070659e-06,
      "loss": 0.5161,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4639
    },
    {
      "epoch": 0.6309491433233614,
      "grad_norm": 0.283203125,
      "learning_rate": 7.210225717203147e-06,
      "loss": 0.623,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4640
    },
    {
      "epoch": 0.6310851237421812,
      "grad_norm": 0.388671875,
      "learning_rate": 7.205666769184867e-06,
      "loss": 0.6972,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4641
    },
    {
      "epoch": 0.6312211041610009,
      "grad_norm": 0.3359375,
      "learning_rate": 7.201108451043461e-06,
      "loss": 0.6403,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4642
    },
    {
      "epoch": 0.6313570845798205,
      "grad_norm": 0.427734375,
      "learning_rate": 7.196550763806429e-06,
      "loss": 0.7297,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4643
    },
    {
      "epoch": 0.6314930649986402,
      "grad_norm": 0.56640625,
      "learning_rate": 7.191993708501131e-06,
      "loss": 0.6242,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4644
    },
    {
      "epoch": 0.6316290454174599,
      "grad_norm": 0.357421875,
      "learning_rate": 7.187437286154782e-06,
      "loss": 0.7313,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4645
    },
    {
      "epoch": 0.6317650258362796,
      "grad_norm": 0.349609375,
      "learning_rate": 7.182881497794455e-06,
      "loss": 0.7371,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4646
    },
    {
      "epoch": 0.6319010062550993,
      "grad_norm": 0.494140625,
      "learning_rate": 7.178326344447081e-06,
      "loss": 0.5394,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4647
    },
    {
      "epoch": 0.6320369866739189,
      "grad_norm": 0.416015625,
      "learning_rate": 7.173771827139446e-06,
      "loss": 0.5667,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4648
    },
    {
      "epoch": 0.6321729670927386,
      "grad_norm": 0.38671875,
      "learning_rate": 7.169217946898197e-06,
      "loss": 0.6419,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4649
    },
    {
      "epoch": 0.6323089475115583,
      "grad_norm": 0.3671875,
      "learning_rate": 7.164664704749831e-06,
      "loss": 0.7826,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4650
    },
    {
      "epoch": 0.632444927930378,
      "grad_norm": 0.447265625,
      "learning_rate": 7.160112101720704e-06,
      "loss": 0.6924,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4651
    },
    {
      "epoch": 0.6325809083491977,
      "grad_norm": 0.30078125,
      "learning_rate": 7.155560138837032e-06,
      "loss": 0.5174,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4652
    },
    {
      "epoch": 0.6327168887680175,
      "grad_norm": 0.5078125,
      "learning_rate": 7.1510088171248805e-06,
      "loss": 0.7021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4653
    },
    {
      "epoch": 0.6328528691868371,
      "grad_norm": 0.578125,
      "learning_rate": 7.146458137610173e-06,
      "loss": 0.7184,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4654
    },
    {
      "epoch": 0.6329888496056568,
      "grad_norm": 0.32421875,
      "learning_rate": 7.141908101318688e-06,
      "loss": 0.6081,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4655
    },
    {
      "epoch": 0.6331248300244765,
      "grad_norm": 0.3515625,
      "learning_rate": 7.137358709276063e-06,
      "loss": 0.651,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4656
    },
    {
      "epoch": 0.6332608104432962,
      "grad_norm": 0.5078125,
      "learning_rate": 7.132809962507783e-06,
      "loss": 0.5958,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4657
    },
    {
      "epoch": 0.6333967908621159,
      "grad_norm": 0.53125,
      "learning_rate": 7.12826186203919e-06,
      "loss": 0.7139,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4658
    },
    {
      "epoch": 0.6335327712809355,
      "grad_norm": 0.361328125,
      "learning_rate": 7.123714408895486e-06,
      "loss": 0.6265,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4659
    },
    {
      "epoch": 0.6336687516997552,
      "grad_norm": 0.416015625,
      "learning_rate": 7.119167604101718e-06,
      "loss": 0.7023,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4660
    },
    {
      "epoch": 0.6338047321185749,
      "grad_norm": 0.435546875,
      "learning_rate": 7.114621448682794e-06,
      "loss": 0.7536,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4661
    },
    {
      "epoch": 0.6339407125373946,
      "grad_norm": 0.48046875,
      "learning_rate": 7.110075943663473e-06,
      "loss": 0.5636,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4662
    },
    {
      "epoch": 0.6340766929562143,
      "grad_norm": 0.32421875,
      "learning_rate": 7.105531090068365e-06,
      "loss": 0.4614,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4663
    },
    {
      "epoch": 0.634212673375034,
      "grad_norm": 0.7265625,
      "learning_rate": 7.100986888921939e-06,
      "loss": 0.821,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4664
    },
    {
      "epoch": 0.6343486537938536,
      "grad_norm": 0.42578125,
      "learning_rate": 7.096443341248511e-06,
      "loss": 0.6657,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4665
    },
    {
      "epoch": 0.6344846342126734,
      "grad_norm": 0.41015625,
      "learning_rate": 7.091900448072253e-06,
      "loss": 0.806,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4666
    },
    {
      "epoch": 0.6346206146314931,
      "grad_norm": 0.49609375,
      "learning_rate": 7.087358210417188e-06,
      "loss": 0.8659,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4667
    },
    {
      "epoch": 0.6347565950503128,
      "grad_norm": 0.291015625,
      "learning_rate": 7.0828166293071945e-06,
      "loss": 0.5523,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4668
    },
    {
      "epoch": 0.6348925754691325,
      "grad_norm": 0.73828125,
      "learning_rate": 7.078275705765996e-06,
      "loss": 0.8856,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4669
    },
    {
      "epoch": 0.6350285558879522,
      "grad_norm": 0.2314453125,
      "learning_rate": 7.073735440817176e-06,
      "loss": 0.4639,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4670
    },
    {
      "epoch": 0.6351645363067718,
      "grad_norm": 0.35546875,
      "learning_rate": 7.069195835484165e-06,
      "loss": 0.651,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4671
    },
    {
      "epoch": 0.6353005167255915,
      "grad_norm": 0.365234375,
      "learning_rate": 7.064656890790245e-06,
      "loss": 0.5407,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4672
    },
    {
      "epoch": 0.6354364971444112,
      "grad_norm": 0.30078125,
      "learning_rate": 7.060118607758548e-06,
      "loss": 0.7121,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4673
    },
    {
      "epoch": 0.6355724775632309,
      "grad_norm": 0.59765625,
      "learning_rate": 7.055580987412061e-06,
      "loss": 0.6685,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4674
    },
    {
      "epoch": 0.6357084579820506,
      "grad_norm": 0.318359375,
      "learning_rate": 7.051044030773619e-06,
      "loss": 0.64,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4675
    },
    {
      "epoch": 0.6358444384008702,
      "grad_norm": 0.349609375,
      "learning_rate": 7.046507738865904e-06,
      "loss": 0.6188,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4676
    },
    {
      "epoch": 0.6359804188196899,
      "grad_norm": 0.40234375,
      "learning_rate": 7.0419721127114545e-06,
      "loss": 0.5496,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4677
    },
    {
      "epoch": 0.6361163992385096,
      "grad_norm": 0.4453125,
      "learning_rate": 7.037437153332658e-06,
      "loss": 0.6203,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4678
    },
    {
      "epoch": 0.6362523796573294,
      "grad_norm": 0.330078125,
      "learning_rate": 7.032902861751747e-06,
      "loss": 0.6604,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4679
    },
    {
      "epoch": 0.6363883600761491,
      "grad_norm": 0.306640625,
      "learning_rate": 7.0283692389908065e-06,
      "loss": 0.5841,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4680
    },
    {
      "epoch": 0.6365243404949688,
      "grad_norm": 0.359375,
      "learning_rate": 7.023836286071771e-06,
      "loss": 0.6351,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4681
    },
    {
      "epoch": 0.6366603209137884,
      "grad_norm": 0.828125,
      "learning_rate": 7.019304004016424e-06,
      "loss": 0.4543,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4682
    },
    {
      "epoch": 0.6367963013326081,
      "grad_norm": 0.3515625,
      "learning_rate": 7.014772393846396e-06,
      "loss": 0.7044,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4683
    },
    {
      "epoch": 0.6369322817514278,
      "grad_norm": 0.392578125,
      "learning_rate": 7.010241456583169e-06,
      "loss": 0.6245,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4684
    },
    {
      "epoch": 0.6370682621702475,
      "grad_norm": 0.416015625,
      "learning_rate": 7.0057111932480705e-06,
      "loss": 0.764,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4685
    },
    {
      "epoch": 0.6372042425890672,
      "grad_norm": 0.494140625,
      "learning_rate": 7.001181604862277e-06,
      "loss": 0.5112,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4686
    },
    {
      "epoch": 0.6373402230078868,
      "grad_norm": 0.5234375,
      "learning_rate": 6.996652692446815e-06,
      "loss": 0.8058,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4687
    },
    {
      "epoch": 0.6374762034267065,
      "grad_norm": 0.380859375,
      "learning_rate": 6.9921244570225536e-06,
      "loss": 0.5443,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4688
    },
    {
      "epoch": 0.6376121838455262,
      "grad_norm": 0.5703125,
      "learning_rate": 6.987596899610214e-06,
      "loss": 0.6579,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4689
    },
    {
      "epoch": 0.6377481642643459,
      "grad_norm": 0.337890625,
      "learning_rate": 6.983070021230364e-06,
      "loss": 0.5916,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4690
    },
    {
      "epoch": 0.6378841446831657,
      "grad_norm": 0.58984375,
      "learning_rate": 6.978543822903417e-06,
      "loss": 0.665,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4691
    },
    {
      "epoch": 0.6380201251019854,
      "grad_norm": 0.39453125,
      "learning_rate": 6.974018305649631e-06,
      "loss": 0.8439,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4692
    },
    {
      "epoch": 0.638156105520805,
      "grad_norm": 0.91015625,
      "learning_rate": 6.969493470489116e-06,
      "loss": 0.9058,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4693
    },
    {
      "epoch": 0.6382920859396247,
      "grad_norm": 0.48046875,
      "learning_rate": 6.9649693184418235e-06,
      "loss": 0.7477,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4694
    },
    {
      "epoch": 0.6384280663584444,
      "grad_norm": 0.369140625,
      "learning_rate": 6.960445850527553e-06,
      "loss": 0.7062,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4695
    },
    {
      "epoch": 0.6385640467772641,
      "grad_norm": 0.5078125,
      "learning_rate": 6.95592306776595e-06,
      "loss": 0.5906,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4696
    },
    {
      "epoch": 0.6387000271960838,
      "grad_norm": 0.546875,
      "learning_rate": 6.951400971176505e-06,
      "loss": 0.6548,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4697
    },
    {
      "epoch": 0.6388360076149034,
      "grad_norm": 0.73828125,
      "learning_rate": 6.946879561778553e-06,
      "loss": 1.0086,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4698
    },
    {
      "epoch": 0.6389719880337231,
      "grad_norm": 0.392578125,
      "learning_rate": 6.942358840591277e-06,
      "loss": 0.522,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4699
    },
    {
      "epoch": 0.6391079684525428,
      "grad_norm": 0.33203125,
      "learning_rate": 6.937838808633699e-06,
      "loss": 0.5929,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4700
    },
    {
      "epoch": 0.6392439488713625,
      "grad_norm": 0.318359375,
      "learning_rate": 6.933319466924693e-06,
      "loss": 0.48,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4701
    },
    {
      "epoch": 0.6393799292901822,
      "grad_norm": 0.71484375,
      "learning_rate": 6.928800816482972e-06,
      "loss": 0.5955,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4702
    },
    {
      "epoch": 0.6395159097090019,
      "grad_norm": 0.60546875,
      "learning_rate": 6.924282858327096e-06,
      "loss": 0.4232,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4703
    },
    {
      "epoch": 0.6396518901278216,
      "grad_norm": 0.369140625,
      "learning_rate": 6.919765593475468e-06,
      "loss": 0.6737,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4704
    },
    {
      "epoch": 0.6397878705466413,
      "grad_norm": 0.328125,
      "learning_rate": 6.9152490229463345e-06,
      "loss": 0.6457,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4705
    },
    {
      "epoch": 0.639923850965461,
      "grad_norm": 0.419921875,
      "learning_rate": 6.910733147757785e-06,
      "loss": 0.617,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4706
    },
    {
      "epoch": 0.6400598313842807,
      "grad_norm": 0.388671875,
      "learning_rate": 6.906217968927754e-06,
      "loss": 0.668,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4707
    },
    {
      "epoch": 0.6401958118031004,
      "grad_norm": 0.369140625,
      "learning_rate": 6.901703487474018e-06,
      "loss": 0.7254,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4708
    },
    {
      "epoch": 0.64033179222192,
      "grad_norm": 0.46875,
      "learning_rate": 6.8971897044141954e-06,
      "loss": 0.577,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4709
    },
    {
      "epoch": 0.6404677726407397,
      "grad_norm": 0.44921875,
      "learning_rate": 6.892676620765748e-06,
      "loss": 0.7472,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4710
    },
    {
      "epoch": 0.6406037530595594,
      "grad_norm": 0.45703125,
      "learning_rate": 6.888164237545982e-06,
      "loss": 0.6961,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4711
    },
    {
      "epoch": 0.6407397334783791,
      "grad_norm": 0.33203125,
      "learning_rate": 6.883652555772042e-06,
      "loss": 0.6317,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4712
    },
    {
      "epoch": 0.6408757138971988,
      "grad_norm": 0.318359375,
      "learning_rate": 6.879141576460917e-06,
      "loss": 0.5041,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4713
    },
    {
      "epoch": 0.6410116943160185,
      "grad_norm": 0.34765625,
      "learning_rate": 6.874631300629435e-06,
      "loss": 0.5547,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4714
    },
    {
      "epoch": 0.6411476747348381,
      "grad_norm": 0.44921875,
      "learning_rate": 6.87012172929427e-06,
      "loss": 0.4612,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4715
    },
    {
      "epoch": 0.6412836551536579,
      "grad_norm": 0.359375,
      "learning_rate": 6.8656128634719334e-06,
      "loss": 0.5729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4716
    },
    {
      "epoch": 0.6414196355724776,
      "grad_norm": 0.33984375,
      "learning_rate": 6.8611047041787785e-06,
      "loss": 0.6559,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4717
    },
    {
      "epoch": 0.6415556159912973,
      "grad_norm": 0.5234375,
      "learning_rate": 6.8565972524309995e-06,
      "loss": 0.8003,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4718
    },
    {
      "epoch": 0.641691596410117,
      "grad_norm": 0.36328125,
      "learning_rate": 6.852090509244631e-06,
      "loss": 0.7116,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4719
    },
    {
      "epoch": 0.6418275768289367,
      "grad_norm": 0.625,
      "learning_rate": 6.847584475635549e-06,
      "loss": 0.7497,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4720
    },
    {
      "epoch": 0.6419635572477563,
      "grad_norm": 0.53515625,
      "learning_rate": 6.843079152619467e-06,
      "loss": 0.6802,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4721
    },
    {
      "epoch": 0.642099537666576,
      "grad_norm": 0.466796875,
      "learning_rate": 6.838574541211942e-06,
      "loss": 0.6567,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4722
    },
    {
      "epoch": 0.6422355180853957,
      "grad_norm": 0.57421875,
      "learning_rate": 6.834070642428368e-06,
      "loss": 0.849,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4723
    },
    {
      "epoch": 0.6423714985042154,
      "grad_norm": 0.259765625,
      "learning_rate": 6.829567457283975e-06,
      "loss": 0.5135,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4724
    },
    {
      "epoch": 0.6425074789230351,
      "grad_norm": 0.255859375,
      "learning_rate": 6.825064986793841e-06,
      "loss": 0.4372,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4725
    },
    {
      "epoch": 0.6426434593418547,
      "grad_norm": 0.314453125,
      "learning_rate": 6.820563231972876e-06,
      "loss": 0.3709,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4726
    },
    {
      "epoch": 0.6427794397606744,
      "grad_norm": 0.498046875,
      "learning_rate": 6.81606219383583e-06,
      "loss": 0.6361,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4727
    },
    {
      "epoch": 0.6429154201794941,
      "grad_norm": 0.19140625,
      "learning_rate": 6.811561873397293e-06,
      "loss": 0.3547,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4728
    },
    {
      "epoch": 0.6430514005983139,
      "grad_norm": 0.37109375,
      "learning_rate": 6.807062271671692e-06,
      "loss": 0.6807,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4729
    },
    {
      "epoch": 0.6431873810171336,
      "grad_norm": 0.296875,
      "learning_rate": 6.802563389673291e-06,
      "loss": 0.4731,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4730
    },
    {
      "epoch": 0.6433233614359533,
      "grad_norm": 0.3828125,
      "learning_rate": 6.798065228416195e-06,
      "loss": 0.6864,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4731
    },
    {
      "epoch": 0.6434593418547729,
      "grad_norm": 0.4375,
      "learning_rate": 6.793567788914343e-06,
      "loss": 0.7303,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4732
    },
    {
      "epoch": 0.6435953222735926,
      "grad_norm": 0.408203125,
      "learning_rate": 6.789071072181516e-06,
      "loss": 0.6878,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4733
    },
    {
      "epoch": 0.6437313026924123,
      "grad_norm": 0.419921875,
      "learning_rate": 6.784575079231324e-06,
      "loss": 0.5929,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4734
    },
    {
      "epoch": 0.643867283111232,
      "grad_norm": 0.431640625,
      "learning_rate": 6.780079811077223e-06,
      "loss": 0.5827,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4735
    },
    {
      "epoch": 0.6440032635300517,
      "grad_norm": 0.384765625,
      "learning_rate": 6.775585268732498e-06,
      "loss": 0.6709,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4736
    },
    {
      "epoch": 0.6441392439488713,
      "grad_norm": 0.330078125,
      "learning_rate": 6.771091453210272e-06,
      "loss": 0.5461,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4737
    },
    {
      "epoch": 0.644275224367691,
      "grad_norm": 0.333984375,
      "learning_rate": 6.766598365523513e-06,
      "loss": 0.6299,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4738
    },
    {
      "epoch": 0.6444112047865107,
      "grad_norm": 0.408203125,
      "learning_rate": 6.762106006685014e-06,
      "loss": 0.7375,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4739
    },
    {
      "epoch": 0.6445471852053304,
      "grad_norm": 0.55078125,
      "learning_rate": 6.757614377707409e-06,
      "loss": 0.6377,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4740
    },
    {
      "epoch": 0.6446831656241501,
      "grad_norm": 0.349609375,
      "learning_rate": 6.753123479603163e-06,
      "loss": 0.6641,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4741
    },
    {
      "epoch": 0.6448191460429699,
      "grad_norm": 1.421875,
      "learning_rate": 6.748633313384583e-06,
      "loss": 0.6631,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4742
    },
    {
      "epoch": 0.6449551264617895,
      "grad_norm": 0.263671875,
      "learning_rate": 6.744143880063805e-06,
      "loss": 0.4547,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4743
    },
    {
      "epoch": 0.6450911068806092,
      "grad_norm": 0.376953125,
      "learning_rate": 6.739655180652802e-06,
      "loss": 0.6706,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4744
    },
    {
      "epoch": 0.6452270872994289,
      "grad_norm": 0.51171875,
      "learning_rate": 6.735167216163383e-06,
      "loss": 0.4295,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4745
    },
    {
      "epoch": 0.6453630677182486,
      "grad_norm": 0.39453125,
      "learning_rate": 6.730679987607189e-06,
      "loss": 0.7867,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4746
    },
    {
      "epoch": 0.6454990481370683,
      "grad_norm": 0.365234375,
      "learning_rate": 6.726193495995699e-06,
      "loss": 0.658,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4747
    },
    {
      "epoch": 0.645635028555888,
      "grad_norm": 0.73046875,
      "learning_rate": 6.7217077423402174e-06,
      "loss": 0.7476,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4748
    },
    {
      "epoch": 0.6457710089747076,
      "grad_norm": 0.50390625,
      "learning_rate": 6.717222727651894e-06,
      "loss": 0.5138,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4749
    },
    {
      "epoch": 0.6459069893935273,
      "grad_norm": 0.3671875,
      "learning_rate": 6.712738452941703e-06,
      "loss": 0.6691,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4750
    },
    {
      "epoch": 0.646042969812347,
      "grad_norm": 0.2890625,
      "learning_rate": 6.708254919220454e-06,
      "loss": 0.3997,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4751
    },
    {
      "epoch": 0.6461789502311667,
      "grad_norm": 0.416015625,
      "learning_rate": 6.703772127498792e-06,
      "loss": 0.7186,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4752
    },
    {
      "epoch": 0.6463149306499864,
      "grad_norm": 0.294921875,
      "learning_rate": 6.699290078787193e-06,
      "loss": 0.5757,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4753
    },
    {
      "epoch": 0.6464509110688061,
      "grad_norm": 0.28515625,
      "learning_rate": 6.6948087740959634e-06,
      "loss": 0.5448,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4754
    },
    {
      "epoch": 0.6465868914876258,
      "grad_norm": 0.2578125,
      "learning_rate": 6.690328214435246e-06,
      "loss": 0.498,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4755
    },
    {
      "epoch": 0.6467228719064455,
      "grad_norm": 0.46875,
      "learning_rate": 6.685848400815014e-06,
      "loss": 0.4811,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4756
    },
    {
      "epoch": 0.6468588523252652,
      "grad_norm": 0.54296875,
      "learning_rate": 6.681369334245071e-06,
      "loss": 0.5679,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4757
    },
    {
      "epoch": 0.6469948327440849,
      "grad_norm": 0.361328125,
      "learning_rate": 6.676891015735055e-06,
      "loss": 0.4999,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4758
    },
    {
      "epoch": 0.6471308131629046,
      "grad_norm": 0.29296875,
      "learning_rate": 6.672413446294433e-06,
      "loss": 0.5783,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4759
    },
    {
      "epoch": 0.6472667935817242,
      "grad_norm": 0.365234375,
      "learning_rate": 6.667936626932502e-06,
      "loss": 0.6118,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4760
    },
    {
      "epoch": 0.6474027740005439,
      "grad_norm": 0.427734375,
      "learning_rate": 6.663460558658394e-06,
      "loss": 0.7093,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4761
    },
    {
      "epoch": 0.6475387544193636,
      "grad_norm": 0.2734375,
      "learning_rate": 6.658985242481069e-06,
      "loss": 0.383,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4762
    },
    {
      "epoch": 0.6476747348381833,
      "grad_norm": 0.5859375,
      "learning_rate": 6.6545106794093175e-06,
      "loss": 0.5941,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4763
    },
    {
      "epoch": 0.647810715257003,
      "grad_norm": 0.181640625,
      "learning_rate": 6.650036870451762e-06,
      "loss": 0.327,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4764
    },
    {
      "epoch": 0.6479466956758226,
      "grad_norm": 0.359375,
      "learning_rate": 6.645563816616851e-06,
      "loss": 0.576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4765
    },
    {
      "epoch": 0.6480826760946423,
      "grad_norm": 0.4140625,
      "learning_rate": 6.6410915189128675e-06,
      "loss": 0.7944,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4766
    },
    {
      "epoch": 0.6482186565134621,
      "grad_norm": 0.76953125,
      "learning_rate": 6.636619978347923e-06,
      "loss": 0.6925,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4767
    },
    {
      "epoch": 0.6483546369322818,
      "grad_norm": 0.2197265625,
      "learning_rate": 6.632149195929957e-06,
      "loss": 0.3659,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4768
    },
    {
      "epoch": 0.6484906173511015,
      "grad_norm": 0.65625,
      "learning_rate": 6.627679172666736e-06,
      "loss": 0.5545,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4769
    },
    {
      "epoch": 0.6486265977699212,
      "grad_norm": 0.244140625,
      "learning_rate": 6.623209909565862e-06,
      "loss": 0.4116,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4770
    },
    {
      "epoch": 0.6487625781887408,
      "grad_norm": 0.478515625,
      "learning_rate": 6.618741407634757e-06,
      "loss": 0.6157,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4771
    },
    {
      "epoch": 0.6488985586075605,
      "grad_norm": 0.40625,
      "learning_rate": 6.6142736678806794e-06,
      "loss": 0.7222,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4772
    },
    {
      "epoch": 0.6490345390263802,
      "grad_norm": 0.373046875,
      "learning_rate": 6.609806691310712e-06,
      "loss": 0.7378,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4773
    },
    {
      "epoch": 0.6491705194451999,
      "grad_norm": 0.28515625,
      "learning_rate": 6.605340478931763e-06,
      "loss": 0.575,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4774
    },
    {
      "epoch": 0.6493064998640196,
      "grad_norm": 0.30859375,
      "learning_rate": 6.600875031750576e-06,
      "loss": 0.7765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4775
    },
    {
      "epoch": 0.6494424802828392,
      "grad_norm": 0.419921875,
      "learning_rate": 6.596410350773712e-06,
      "loss": 0.6117,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4776
    },
    {
      "epoch": 0.6495784607016589,
      "grad_norm": 0.37109375,
      "learning_rate": 6.591946437007569e-06,
      "loss": 0.6296,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4777
    },
    {
      "epoch": 0.6497144411204786,
      "grad_norm": 0.359375,
      "learning_rate": 6.5874832914583665e-06,
      "loss": 0.5269,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4778
    },
    {
      "epoch": 0.6498504215392984,
      "grad_norm": 0.41796875,
      "learning_rate": 6.5830209151321525e-06,
      "loss": 0.7046,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4779
    },
    {
      "epoch": 0.6499864019581181,
      "grad_norm": 0.328125,
      "learning_rate": 6.5785593090347975e-06,
      "loss": 0.6019,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4780
    },
    {
      "epoch": 0.6501223823769378,
      "grad_norm": 0.396484375,
      "learning_rate": 6.574098474172008e-06,
      "loss": 0.5622,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4781
    },
    {
      "epoch": 0.6502583627957574,
      "grad_norm": 0.2734375,
      "learning_rate": 6.569638411549306e-06,
      "loss": 0.45,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4782
    },
    {
      "epoch": 0.6503943432145771,
      "grad_norm": 0.48046875,
      "learning_rate": 6.565179122172047e-06,
      "loss": 0.8919,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4783
    },
    {
      "epoch": 0.6505303236333968,
      "grad_norm": 0.283203125,
      "learning_rate": 6.560720607045407e-06,
      "loss": 0.4865,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4784
    },
    {
      "epoch": 0.6506663040522165,
      "grad_norm": 0.451171875,
      "learning_rate": 6.556262867174393e-06,
      "loss": 0.6706,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4785
    },
    {
      "epoch": 0.6508022844710362,
      "grad_norm": 0.546875,
      "learning_rate": 6.551805903563831e-06,
      "loss": 0.4225,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4786
    },
    {
      "epoch": 0.6509382648898558,
      "grad_norm": 0.302734375,
      "learning_rate": 6.547349717218377e-06,
      "loss": 0.4751,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4787
    },
    {
      "epoch": 0.6510742453086755,
      "grad_norm": 1.1171875,
      "learning_rate": 6.542894309142509e-06,
      "loss": 0.7121,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4788
    },
    {
      "epoch": 0.6512102257274952,
      "grad_norm": 0.5390625,
      "learning_rate": 6.538439680340532e-06,
      "loss": 0.7612,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4789
    },
    {
      "epoch": 0.6513462061463149,
      "grad_norm": 0.3671875,
      "learning_rate": 6.533985831816571e-06,
      "loss": 0.5889,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4790
    },
    {
      "epoch": 0.6514821865651346,
      "grad_norm": 0.369140625,
      "learning_rate": 6.529532764574578e-06,
      "loss": 0.7453,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4791
    },
    {
      "epoch": 0.6516181669839544,
      "grad_norm": 0.341796875,
      "learning_rate": 6.525080479618331e-06,
      "loss": 0.6668,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4792
    },
    {
      "epoch": 0.651754147402774,
      "grad_norm": 0.62890625,
      "learning_rate": 6.520628977951428e-06,
      "loss": 0.7782,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4793
    },
    {
      "epoch": 0.6518901278215937,
      "grad_norm": 0.384765625,
      "learning_rate": 6.516178260577293e-06,
      "loss": 0.7059,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4794
    },
    {
      "epoch": 0.6520261082404134,
      "grad_norm": 0.314453125,
      "learning_rate": 6.511728328499172e-06,
      "loss": 0.6147,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4795
    },
    {
      "epoch": 0.6521620886592331,
      "grad_norm": 0.59375,
      "learning_rate": 6.507279182720132e-06,
      "loss": 0.4567,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4796
    },
    {
      "epoch": 0.6522980690780528,
      "grad_norm": 0.298828125,
      "learning_rate": 6.5028308242430635e-06,
      "loss": 0.651,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4797
    },
    {
      "epoch": 0.6524340494968724,
      "grad_norm": 0.232421875,
      "learning_rate": 6.498383254070688e-06,
      "loss": 0.4531,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4798
    },
    {
      "epoch": 0.6525700299156921,
      "grad_norm": 0.30078125,
      "learning_rate": 6.493936473205538e-06,
      "loss": 0.4617,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4799
    },
    {
      "epoch": 0.6527060103345118,
      "grad_norm": 0.466796875,
      "learning_rate": 6.489490482649972e-06,
      "loss": 0.6908,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4800
    },
    {
      "epoch": 0.6528419907533315,
      "grad_norm": 0.337890625,
      "learning_rate": 6.4850452834061716e-06,
      "loss": 0.6533,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4801
    },
    {
      "epoch": 0.6529779711721512,
      "grad_norm": 0.333984375,
      "learning_rate": 6.480600876476137e-06,
      "loss": 0.7132,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4802
    },
    {
      "epoch": 0.6531139515909709,
      "grad_norm": 0.40234375,
      "learning_rate": 6.476157262861695e-06,
      "loss": 0.6359,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4803
    },
    {
      "epoch": 0.6532499320097905,
      "grad_norm": 0.546875,
      "learning_rate": 6.471714443564488e-06,
      "loss": 0.5122,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4804
    },
    {
      "epoch": 0.6533859124286103,
      "grad_norm": 0.2451171875,
      "learning_rate": 6.467272419585984e-06,
      "loss": 0.388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4805
    },
    {
      "epoch": 0.65352189284743,
      "grad_norm": 0.345703125,
      "learning_rate": 6.462831191927466e-06,
      "loss": 0.6466,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4806
    },
    {
      "epoch": 0.6536578732662497,
      "grad_norm": 0.251953125,
      "learning_rate": 6.458390761590043e-06,
      "loss": 0.445,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4807
    },
    {
      "epoch": 0.6537938536850694,
      "grad_norm": 0.431640625,
      "learning_rate": 6.453951129574644e-06,
      "loss": 0.7588,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4808
    },
    {
      "epoch": 0.653929834103889,
      "grad_norm": 0.3359375,
      "learning_rate": 6.449512296882013e-06,
      "loss": 0.5968,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4809
    },
    {
      "epoch": 0.6540658145227087,
      "grad_norm": 0.484375,
      "learning_rate": 6.445074264512721e-06,
      "loss": 0.7793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4810
    },
    {
      "epoch": 0.6542017949415284,
      "grad_norm": 0.2041015625,
      "learning_rate": 6.4406370334671495e-06,
      "loss": 0.3737,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4811
    },
    {
      "epoch": 0.6543377753603481,
      "grad_norm": 0.50390625,
      "learning_rate": 6.4362006047455105e-06,
      "loss": 0.8027,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4812
    },
    {
      "epoch": 0.6544737557791678,
      "grad_norm": 0.42578125,
      "learning_rate": 6.431764979347826e-06,
      "loss": 0.6615,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4813
    },
    {
      "epoch": 0.6546097361979875,
      "grad_norm": 0.494140625,
      "learning_rate": 6.427330158273939e-06,
      "loss": 0.5493,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4814
    },
    {
      "epoch": 0.6547457166168071,
      "grad_norm": 0.416015625,
      "learning_rate": 6.422896142523515e-06,
      "loss": 0.6929,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4815
    },
    {
      "epoch": 0.6548816970356268,
      "grad_norm": 0.33984375,
      "learning_rate": 6.418462933096035e-06,
      "loss": 0.5399,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4816
    },
    {
      "epoch": 0.6550176774544466,
      "grad_norm": 0.373046875,
      "learning_rate": 6.414030530990797e-06,
      "loss": 0.6748,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4817
    },
    {
      "epoch": 0.6551536578732663,
      "grad_norm": 0.36328125,
      "learning_rate": 6.40959893720692e-06,
      "loss": 0.7388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4818
    },
    {
      "epoch": 0.655289638292086,
      "grad_norm": 0.412109375,
      "learning_rate": 6.40516815274334e-06,
      "loss": 0.5765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4819
    },
    {
      "epoch": 0.6554256187109057,
      "grad_norm": 0.3671875,
      "learning_rate": 6.400738178598809e-06,
      "loss": 0.6082,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4820
    },
    {
      "epoch": 0.6555615991297253,
      "grad_norm": 0.3125,
      "learning_rate": 6.3963090157718955e-06,
      "loss": 0.5355,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4821
    },
    {
      "epoch": 0.655697579548545,
      "grad_norm": 0.322265625,
      "learning_rate": 6.39188066526099e-06,
      "loss": 0.5571,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4822
    },
    {
      "epoch": 0.6558335599673647,
      "grad_norm": 0.48828125,
      "learning_rate": 6.387453128064296e-06,
      "loss": 0.7067,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4823
    },
    {
      "epoch": 0.6559695403861844,
      "grad_norm": 0.294921875,
      "learning_rate": 6.383026405179835e-06,
      "loss": 0.4677,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4824
    },
    {
      "epoch": 0.6561055208050041,
      "grad_norm": 0.380859375,
      "learning_rate": 6.378600497605443e-06,
      "loss": 0.5218,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4825
    },
    {
      "epoch": 0.6562415012238237,
      "grad_norm": 0.60546875,
      "learning_rate": 6.374175406338774e-06,
      "loss": 0.5454,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4826
    },
    {
      "epoch": 0.6563774816426434,
      "grad_norm": 0.75,
      "learning_rate": 6.369751132377298e-06,
      "loss": 0.7985,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4827
    },
    {
      "epoch": 0.6565134620614631,
      "grad_norm": 0.287109375,
      "learning_rate": 6.3653276767183e-06,
      "loss": 0.5529,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4828
    },
    {
      "epoch": 0.6566494424802828,
      "grad_norm": 0.54296875,
      "learning_rate": 6.360905040358882e-06,
      "loss": 0.6746,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4829
    },
    {
      "epoch": 0.6567854228991026,
      "grad_norm": 0.8515625,
      "learning_rate": 6.356483224295959e-06,
      "loss": 0.8034,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4830
    },
    {
      "epoch": 0.6569214033179223,
      "grad_norm": 0.259765625,
      "learning_rate": 6.352062229526266e-06,
      "loss": 0.5218,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4831
    },
    {
      "epoch": 0.6570573837367419,
      "grad_norm": 0.298828125,
      "learning_rate": 6.347642057046344e-06,
      "loss": 0.3908,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4832
    },
    {
      "epoch": 0.6571933641555616,
      "grad_norm": 1.59375,
      "learning_rate": 6.3432227078525564e-06,
      "loss": 0.8179,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4833
    },
    {
      "epoch": 0.6573293445743813,
      "grad_norm": 0.408203125,
      "learning_rate": 6.3388041829410785e-06,
      "loss": 0.7279,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4834
    },
    {
      "epoch": 0.657465324993201,
      "grad_norm": 0.859375,
      "learning_rate": 6.334386483307901e-06,
      "loss": 0.7983,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4835
    },
    {
      "epoch": 0.6576013054120207,
      "grad_norm": 0.453125,
      "learning_rate": 6.329969609948825e-06,
      "loss": 0.7651,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4836
    },
    {
      "epoch": 0.6577372858308403,
      "grad_norm": 0.31640625,
      "learning_rate": 6.32555356385947e-06,
      "loss": 0.4694,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4837
    },
    {
      "epoch": 0.65787326624966,
      "grad_norm": 0.353515625,
      "learning_rate": 6.321138346035265e-06,
      "loss": 0.5612,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4838
    },
    {
      "epoch": 0.6580092466684797,
      "grad_norm": 0.6953125,
      "learning_rate": 6.3167239574714556e-06,
      "loss": 0.7673,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4839
    },
    {
      "epoch": 0.6581452270872994,
      "grad_norm": 0.53515625,
      "learning_rate": 6.312310399163098e-06,
      "loss": 0.7388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4840
    },
    {
      "epoch": 0.6582812075061191,
      "grad_norm": 0.640625,
      "learning_rate": 6.307897672105061e-06,
      "loss": 0.7087,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4841
    },
    {
      "epoch": 0.6584171879249389,
      "grad_norm": 0.412109375,
      "learning_rate": 6.303485777292031e-06,
      "loss": 0.5347,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4842
    },
    {
      "epoch": 0.6585531683437585,
      "grad_norm": 0.478515625,
      "learning_rate": 6.2990747157185e-06,
      "loss": 0.7664,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4843
    },
    {
      "epoch": 0.6586891487625782,
      "grad_norm": 0.58984375,
      "learning_rate": 6.294664488378777e-06,
      "loss": 0.589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4844
    },
    {
      "epoch": 0.6588251291813979,
      "grad_norm": 0.32421875,
      "learning_rate": 6.2902550962669775e-06,
      "loss": 0.5741,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4845
    },
    {
      "epoch": 0.6589611096002176,
      "grad_norm": 0.326171875,
      "learning_rate": 6.285846540377036e-06,
      "loss": 0.6017,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4846
    },
    {
      "epoch": 0.6590970900190373,
      "grad_norm": 0.298828125,
      "learning_rate": 6.281438821702694e-06,
      "loss": 0.6062,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4847
    },
    {
      "epoch": 0.659233070437857,
      "grad_norm": 0.279296875,
      "learning_rate": 6.277031941237504e-06,
      "loss": 0.561,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4848
    },
    {
      "epoch": 0.6593690508566766,
      "grad_norm": 0.81640625,
      "learning_rate": 6.272625899974833e-06,
      "loss": 0.4759,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4849
    },
    {
      "epoch": 0.6595050312754963,
      "grad_norm": 0.357421875,
      "learning_rate": 6.268220698907854e-06,
      "loss": 0.6374,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4850
    },
    {
      "epoch": 0.659641011694316,
      "grad_norm": 0.5859375,
      "learning_rate": 6.263816339029556e-06,
      "loss": 0.8309,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4851
    },
    {
      "epoch": 0.6597769921131357,
      "grad_norm": 0.2734375,
      "learning_rate": 6.2594128213327336e-06,
      "loss": 0.5018,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4852
    },
    {
      "epoch": 0.6599129725319554,
      "grad_norm": 0.251953125,
      "learning_rate": 6.2550101468099935e-06,
      "loss": 0.4648,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4853
    },
    {
      "epoch": 0.660048952950775,
      "grad_norm": 0.5859375,
      "learning_rate": 6.250608316453752e-06,
      "loss": 0.7887,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4854
    },
    {
      "epoch": 0.6601849333695948,
      "grad_norm": 0.365234375,
      "learning_rate": 6.246207331256238e-06,
      "loss": 0.658,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4855
    },
    {
      "epoch": 0.6603209137884145,
      "grad_norm": 0.455078125,
      "learning_rate": 6.241807192209486e-06,
      "loss": 0.7819,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4856
    },
    {
      "epoch": 0.6604568942072342,
      "grad_norm": 0.353515625,
      "learning_rate": 6.237407900305334e-06,
      "loss": 0.6724,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4857
    },
    {
      "epoch": 0.6605928746260539,
      "grad_norm": 0.35546875,
      "learning_rate": 6.23300945653545e-06,
      "loss": 0.832,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4858
    },
    {
      "epoch": 0.6607288550448736,
      "grad_norm": 0.228515625,
      "learning_rate": 6.228611861891289e-06,
      "loss": 0.4567,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4859
    },
    {
      "epoch": 0.6608648354636932,
      "grad_norm": 0.322265625,
      "learning_rate": 6.224215117364124e-06,
      "loss": 0.6561,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4860
    },
    {
      "epoch": 0.6610008158825129,
      "grad_norm": 0.3515625,
      "learning_rate": 6.219819223945034e-06,
      "loss": 0.5793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4861
    },
    {
      "epoch": 0.6611367963013326,
      "grad_norm": 0.296875,
      "learning_rate": 6.2154241826249075e-06,
      "loss": 0.5452,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4862
    },
    {
      "epoch": 0.6612727767201523,
      "grad_norm": 0.2490234375,
      "learning_rate": 6.211029994394442e-06,
      "loss": 0.4357,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4863
    },
    {
      "epoch": 0.661408757138972,
      "grad_norm": 0.5078125,
      "learning_rate": 6.206636660244138e-06,
      "loss": 0.5861,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4864
    },
    {
      "epoch": 0.6615447375577916,
      "grad_norm": 0.3203125,
      "learning_rate": 6.202244181164311e-06,
      "loss": 0.5617,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4865
    },
    {
      "epoch": 0.6616807179766113,
      "grad_norm": 0.859375,
      "learning_rate": 6.197852558145077e-06,
      "loss": 0.8835,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4866
    },
    {
      "epoch": 0.6618166983954311,
      "grad_norm": 0.5625,
      "learning_rate": 6.19346179217636e-06,
      "loss": 0.7407,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4867
    },
    {
      "epoch": 0.6619526788142508,
      "grad_norm": 0.2734375,
      "learning_rate": 6.189071884247895e-06,
      "loss": 0.3669,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4868
    },
    {
      "epoch": 0.6620886592330705,
      "grad_norm": 0.263671875,
      "learning_rate": 6.184682835349221e-06,
      "loss": 0.5249,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4869
    },
    {
      "epoch": 0.6622246396518902,
      "grad_norm": 0.193359375,
      "learning_rate": 6.18029464646968e-06,
      "loss": 0.3667,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4870
    },
    {
      "epoch": 0.6623606200707098,
      "grad_norm": 0.359375,
      "learning_rate": 6.175907318598427e-06,
      "loss": 0.508,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4871
    },
    {
      "epoch": 0.6624966004895295,
      "grad_norm": 0.39453125,
      "learning_rate": 6.171520852724419e-06,
      "loss": 0.7036,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4872
    },
    {
      "epoch": 0.6626325809083492,
      "grad_norm": 0.376953125,
      "learning_rate": 6.167135249836417e-06,
      "loss": 0.5392,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4873
    },
    {
      "epoch": 0.6627685613271689,
      "grad_norm": 0.54296875,
      "learning_rate": 6.1627505109229906e-06,
      "loss": 0.591,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4874
    },
    {
      "epoch": 0.6629045417459886,
      "grad_norm": 0.400390625,
      "learning_rate": 6.158366636972512e-06,
      "loss": 0.6525,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4875
    },
    {
      "epoch": 0.6630405221648082,
      "grad_norm": 0.515625,
      "learning_rate": 6.153983628973165e-06,
      "loss": 0.6003,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4876
    },
    {
      "epoch": 0.6631765025836279,
      "grad_norm": 0.6875,
      "learning_rate": 6.149601487912927e-06,
      "loss": 0.744,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4877
    },
    {
      "epoch": 0.6633124830024476,
      "grad_norm": 0.318359375,
      "learning_rate": 6.14522021477959e-06,
      "loss": 0.6126,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4878
    },
    {
      "epoch": 0.6634484634212673,
      "grad_norm": 0.34375,
      "learning_rate": 6.140839810560746e-06,
      "loss": 0.6164,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4879
    },
    {
      "epoch": 0.6635844438400871,
      "grad_norm": 0.373046875,
      "learning_rate": 6.1364602762437896e-06,
      "loss": 0.77,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4880
    },
    {
      "epoch": 0.6637204242589068,
      "grad_norm": 0.35546875,
      "learning_rate": 6.132081612815925e-06,
      "loss": 0.6188,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4881
    },
    {
      "epoch": 0.6638564046777264,
      "grad_norm": 0.392578125,
      "learning_rate": 6.127703821264152e-06,
      "loss": 0.7092,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4882
    },
    {
      "epoch": 0.6639923850965461,
      "grad_norm": 0.291015625,
      "learning_rate": 6.123326902575282e-06,
      "loss": 0.6528,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4883
    },
    {
      "epoch": 0.6641283655153658,
      "grad_norm": 2.046875,
      "learning_rate": 6.118950857735924e-06,
      "loss": 0.6543,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4884
    },
    {
      "epoch": 0.6642643459341855,
      "grad_norm": 0.734375,
      "learning_rate": 6.114575687732493e-06,
      "loss": 0.6035,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4885
    },
    {
      "epoch": 0.6644003263530052,
      "grad_norm": 0.34765625,
      "learning_rate": 6.1102013935512065e-06,
      "loss": 0.5524,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4886
    },
    {
      "epoch": 0.6645363067718248,
      "grad_norm": 0.578125,
      "learning_rate": 6.105827976178082e-06,
      "loss": 0.734,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4887
    },
    {
      "epoch": 0.6646722871906445,
      "grad_norm": 0.375,
      "learning_rate": 6.101455436598942e-06,
      "loss": 0.6056,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4888
    },
    {
      "epoch": 0.6648082676094642,
      "grad_norm": 0.29296875,
      "learning_rate": 6.097083775799411e-06,
      "loss": 0.4987,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4889
    },
    {
      "epoch": 0.6649442480282839,
      "grad_norm": 0.29296875,
      "learning_rate": 6.0927129947649136e-06,
      "loss": 0.5747,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4890
    },
    {
      "epoch": 0.6650802284471036,
      "grad_norm": 0.34375,
      "learning_rate": 6.088343094480677e-06,
      "loss": 0.591,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4891
    },
    {
      "epoch": 0.6652162088659233,
      "grad_norm": 0.73828125,
      "learning_rate": 6.0839740759317335e-06,
      "loss": 0.7253,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4892
    },
    {
      "epoch": 0.665352189284743,
      "grad_norm": 0.3046875,
      "learning_rate": 6.0796059401029096e-06,
      "loss": 0.5365,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4893
    },
    {
      "epoch": 0.6654881697035627,
      "grad_norm": 0.6796875,
      "learning_rate": 6.075238687978837e-06,
      "loss": 0.682,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4894
    },
    {
      "epoch": 0.6656241501223824,
      "grad_norm": 0.435546875,
      "learning_rate": 6.07087232054395e-06,
      "loss": 0.7404,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4895
    },
    {
      "epoch": 0.6657601305412021,
      "grad_norm": 0.52734375,
      "learning_rate": 6.06650683878248e-06,
      "loss": 0.8812,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4896
    },
    {
      "epoch": 0.6658961109600218,
      "grad_norm": 0.296875,
      "learning_rate": 6.062142243678459e-06,
      "loss": 0.605,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4897
    },
    {
      "epoch": 0.6660320913788415,
      "grad_norm": 0.484375,
      "learning_rate": 6.057778536215724e-06,
      "loss": 0.7204,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4898
    },
    {
      "epoch": 0.6661680717976611,
      "grad_norm": 0.5234375,
      "learning_rate": 6.053415717377902e-06,
      "loss": 0.6143,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4899
    },
    {
      "epoch": 0.6663040522164808,
      "grad_norm": 0.208984375,
      "learning_rate": 6.049053788148427e-06,
      "loss": 0.3325,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4900
    },
    {
      "epoch": 0.6664400326353005,
      "grad_norm": 0.25390625,
      "learning_rate": 6.0446927495105345e-06,
      "loss": 0.4701,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4901
    },
    {
      "epoch": 0.6665760130541202,
      "grad_norm": 0.45703125,
      "learning_rate": 6.0403326024472545e-06,
      "loss": 0.4627,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4902
    },
    {
      "epoch": 0.6667119934729399,
      "grad_norm": 0.333984375,
      "learning_rate": 6.0359733479414164e-06,
      "loss": 0.6356,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4903
    },
    {
      "epoch": 0.6668479738917595,
      "grad_norm": 0.33984375,
      "learning_rate": 6.031614986975649e-06,
      "loss": 0.6437,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4904
    },
    {
      "epoch": 0.6669839543105793,
      "grad_norm": 0.306640625,
      "learning_rate": 6.027257520532384e-06,
      "loss": 0.5228,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4905
    },
    {
      "epoch": 0.667119934729399,
      "grad_norm": 0.482421875,
      "learning_rate": 6.022900949593843e-06,
      "loss": 0.3003,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4906
    },
    {
      "epoch": 0.6672559151482187,
      "grad_norm": 0.369140625,
      "learning_rate": 6.018545275142053e-06,
      "loss": 0.7155,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4907
    },
    {
      "epoch": 0.6673918955670384,
      "grad_norm": 0.6953125,
      "learning_rate": 6.014190498158835e-06,
      "loss": 0.7048,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4908
    },
    {
      "epoch": 0.667527875985858,
      "grad_norm": 0.427734375,
      "learning_rate": 6.0098366196258095e-06,
      "loss": 0.7568,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4909
    },
    {
      "epoch": 0.6676638564046777,
      "grad_norm": 0.298828125,
      "learning_rate": 6.005483640524394e-06,
      "loss": 0.4826,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4910
    },
    {
      "epoch": 0.6677998368234974,
      "grad_norm": 0.421875,
      "learning_rate": 6.001131561835804e-06,
      "loss": 0.8397,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4911
    },
    {
      "epoch": 0.6679358172423171,
      "grad_norm": 0.5625,
      "learning_rate": 5.99678038454105e-06,
      "loss": 0.6014,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4912
    },
    {
      "epoch": 0.6680717976611368,
      "grad_norm": 0.49609375,
      "learning_rate": 5.9924301096209415e-06,
      "loss": 0.695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4913
    },
    {
      "epoch": 0.6682077780799565,
      "grad_norm": 0.59765625,
      "learning_rate": 5.988080738056083e-06,
      "loss": 0.6475,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4914
    },
    {
      "epoch": 0.6683437584987761,
      "grad_norm": 0.52734375,
      "learning_rate": 5.983732270826876e-06,
      "loss": 0.6201,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4915
    },
    {
      "epoch": 0.6684797389175958,
      "grad_norm": 0.5703125,
      "learning_rate": 5.979384708913518e-06,
      "loss": 0.7826,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4916
    },
    {
      "epoch": 0.6686157193364155,
      "grad_norm": 0.421875,
      "learning_rate": 5.9750380532959985e-06,
      "loss": 0.6432,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4917
    },
    {
      "epoch": 0.6687516997552353,
      "grad_norm": 0.34375,
      "learning_rate": 5.970692304954115e-06,
      "loss": 0.7039,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4918
    },
    {
      "epoch": 0.668887680174055,
      "grad_norm": 0.400390625,
      "learning_rate": 5.96634746486745e-06,
      "loss": 0.79,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4919
    },
    {
      "epoch": 0.6690236605928747,
      "grad_norm": 0.345703125,
      "learning_rate": 5.962003534015382e-06,
      "loss": 0.7205,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4920
    },
    {
      "epoch": 0.6691596410116943,
      "grad_norm": 0.37890625,
      "learning_rate": 5.957660513377086e-06,
      "loss": 0.6447,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4921
    },
    {
      "epoch": 0.669295621430514,
      "grad_norm": 0.35546875,
      "learning_rate": 5.953318403931533e-06,
      "loss": 0.7425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4922
    },
    {
      "epoch": 0.6694316018493337,
      "grad_norm": 0.302734375,
      "learning_rate": 5.9489772066574854e-06,
      "loss": 0.467,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4923
    },
    {
      "epoch": 0.6695675822681534,
      "grad_norm": 0.341796875,
      "learning_rate": 5.944636922533504e-06,
      "loss": 0.7695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4924
    },
    {
      "epoch": 0.6697035626869731,
      "grad_norm": 0.2734375,
      "learning_rate": 5.940297552537942e-06,
      "loss": 0.4655,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4925
    },
    {
      "epoch": 0.6698395431057927,
      "grad_norm": 0.30859375,
      "learning_rate": 5.935959097648946e-06,
      "loss": 0.5815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4926
    },
    {
      "epoch": 0.6699755235246124,
      "grad_norm": 0.474609375,
      "learning_rate": 5.931621558844456e-06,
      "loss": 0.5749,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4927
    },
    {
      "epoch": 0.6701115039434321,
      "grad_norm": 0.375,
      "learning_rate": 5.927284937102208e-06,
      "loss": 0.6913,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4928
    },
    {
      "epoch": 0.6702474843622518,
      "grad_norm": 0.51171875,
      "learning_rate": 5.922949233399727e-06,
      "loss": 0.4642,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4929
    },
    {
      "epoch": 0.6703834647810716,
      "grad_norm": 0.9140625,
      "learning_rate": 5.918614448714337e-06,
      "loss": 0.8211,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4930
    },
    {
      "epoch": 0.6705194451998913,
      "grad_norm": 0.7578125,
      "learning_rate": 5.914280584023149e-06,
      "loss": 0.673,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4931
    },
    {
      "epoch": 0.6706554256187109,
      "grad_norm": 0.486328125,
      "learning_rate": 5.909947640303068e-06,
      "loss": 0.7531,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4932
    },
    {
      "epoch": 0.6707914060375306,
      "grad_norm": 0.38671875,
      "learning_rate": 5.905615618530795e-06,
      "loss": 0.6486,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4933
    },
    {
      "epoch": 0.6709273864563503,
      "grad_norm": 0.31640625,
      "learning_rate": 5.90128451968282e-06,
      "loss": 0.5866,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4934
    },
    {
      "epoch": 0.67106336687517,
      "grad_norm": 0.412109375,
      "learning_rate": 5.896954344735426e-06,
      "loss": 0.6916,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4935
    },
    {
      "epoch": 0.6711993472939897,
      "grad_norm": 0.390625,
      "learning_rate": 5.892625094664688e-06,
      "loss": 0.5776,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4936
    },
    {
      "epoch": 0.6713353277128093,
      "grad_norm": 0.3984375,
      "learning_rate": 5.8882967704464685e-06,
      "loss": 0.6444,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4937
    },
    {
      "epoch": 0.671471308131629,
      "grad_norm": 0.40234375,
      "learning_rate": 5.883969373056428e-06,
      "loss": 0.6587,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4938
    },
    {
      "epoch": 0.6716072885504487,
      "grad_norm": 0.3828125,
      "learning_rate": 5.879642903470013e-06,
      "loss": 0.5809,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4939
    },
    {
      "epoch": 0.6717432689692684,
      "grad_norm": 0.6015625,
      "learning_rate": 5.875317362662465e-06,
      "loss": 0.7805,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4940
    },
    {
      "epoch": 0.6718792493880881,
      "grad_norm": 0.390625,
      "learning_rate": 5.870992751608811e-06,
      "loss": 0.741,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4941
    },
    {
      "epoch": 0.6720152298069078,
      "grad_norm": 0.302734375,
      "learning_rate": 5.8666690712838715e-06,
      "loss": 0.4857,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4942
    },
    {
      "epoch": 0.6721512102257275,
      "grad_norm": 0.412109375,
      "learning_rate": 5.862346322662259e-06,
      "loss": 0.7294,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4943
    },
    {
      "epoch": 0.6722871906445472,
      "grad_norm": 0.33984375,
      "learning_rate": 5.8580245067183714e-06,
      "loss": 0.765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4944
    },
    {
      "epoch": 0.6724231710633669,
      "grad_norm": 1.3984375,
      "learning_rate": 5.8537036244264014e-06,
      "loss": 0.9166,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4945
    },
    {
      "epoch": 0.6725591514821866,
      "grad_norm": 0.30078125,
      "learning_rate": 5.849383676760327e-06,
      "loss": 0.6538,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4946
    },
    {
      "epoch": 0.6726951319010063,
      "grad_norm": 0.52734375,
      "learning_rate": 5.845064664693918e-06,
      "loss": 0.7096,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4947
    },
    {
      "epoch": 0.672831112319826,
      "grad_norm": 0.2333984375,
      "learning_rate": 5.840746589200732e-06,
      "loss": 0.4499,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4948
    },
    {
      "epoch": 0.6729670927386456,
      "grad_norm": 0.326171875,
      "learning_rate": 5.8364294512541174e-06,
      "loss": 0.5487,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4949
    },
    {
      "epoch": 0.6731030731574653,
      "grad_norm": 0.4453125,
      "learning_rate": 5.832113251827205e-06,
      "loss": 0.667,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4950
    },
    {
      "epoch": 0.673239053576285,
      "grad_norm": 0.48828125,
      "learning_rate": 5.827797991892927e-06,
      "loss": 0.8198,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4951
    },
    {
      "epoch": 0.6733750339951047,
      "grad_norm": 0.56640625,
      "learning_rate": 5.823483672423988e-06,
      "loss": 0.5827,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4952
    },
    {
      "epoch": 0.6735110144139244,
      "grad_norm": 0.291015625,
      "learning_rate": 5.819170294392894e-06,
      "loss": 0.6129,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4953
    },
    {
      "epoch": 0.673646994832744,
      "grad_norm": 0.72265625,
      "learning_rate": 5.814857858771928e-06,
      "loss": 0.8281,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4954
    },
    {
      "epoch": 0.6737829752515637,
      "grad_norm": 0.34765625,
      "learning_rate": 5.810546366533172e-06,
      "loss": 0.6632,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4955
    },
    {
      "epoch": 0.6739189556703835,
      "grad_norm": 0.4296875,
      "learning_rate": 5.806235818648481e-06,
      "loss": 0.7153,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4956
    },
    {
      "epoch": 0.6740549360892032,
      "grad_norm": 0.357421875,
      "learning_rate": 5.801926216089515e-06,
      "loss": 0.4519,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4957
    },
    {
      "epoch": 0.6741909165080229,
      "grad_norm": 0.462890625,
      "learning_rate": 5.797617559827702e-06,
      "loss": 0.7746,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4958
    },
    {
      "epoch": 0.6743268969268426,
      "grad_norm": 0.326171875,
      "learning_rate": 5.793309850834272e-06,
      "loss": 0.5197,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4959
    },
    {
      "epoch": 0.6744628773456622,
      "grad_norm": 0.44140625,
      "learning_rate": 5.78900309008023e-06,
      "loss": 0.6061,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4960
    },
    {
      "epoch": 0.6745988577644819,
      "grad_norm": 0.609375,
      "learning_rate": 5.78469727853638e-06,
      "loss": 0.8729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4961
    },
    {
      "epoch": 0.6747348381833016,
      "grad_norm": 0.63671875,
      "learning_rate": 5.780392417173295e-06,
      "loss": 0.7661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4962
    },
    {
      "epoch": 0.6748708186021213,
      "grad_norm": 0.4453125,
      "learning_rate": 5.776088506961353e-06,
      "loss": 0.6559,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4963
    },
    {
      "epoch": 0.675006799020941,
      "grad_norm": 0.400390625,
      "learning_rate": 5.771785548870698e-06,
      "loss": 0.7741,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4964
    },
    {
      "epoch": 0.6751427794397606,
      "grad_norm": 0.39453125,
      "learning_rate": 5.7674835438712805e-06,
      "loss": 0.7997,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4965
    },
    {
      "epoch": 0.6752787598585803,
      "grad_norm": 0.345703125,
      "learning_rate": 5.763182492932815e-06,
      "loss": 0.5941,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4966
    },
    {
      "epoch": 0.6754147402774,
      "grad_norm": 0.3046875,
      "learning_rate": 5.7588823970248185e-06,
      "loss": 0.5462,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4967
    },
    {
      "epoch": 0.6755507206962198,
      "grad_norm": 0.376953125,
      "learning_rate": 5.754583257116579e-06,
      "loss": 0.8013,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4968
    },
    {
      "epoch": 0.6756867011150395,
      "grad_norm": 0.251953125,
      "learning_rate": 5.750285074177182e-06,
      "loss": 0.5106,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4969
    },
    {
      "epoch": 0.6758226815338592,
      "grad_norm": 0.2890625,
      "learning_rate": 5.745987849175483e-06,
      "loss": 0.6151,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4970
    },
    {
      "epoch": 0.6759586619526788,
      "grad_norm": 0.40234375,
      "learning_rate": 5.741691583080135e-06,
      "loss": 0.6287,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4971
    },
    {
      "epoch": 0.6760946423714985,
      "grad_norm": 0.3515625,
      "learning_rate": 5.737396276859564e-06,
      "loss": 0.4798,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4972
    },
    {
      "epoch": 0.6762306227903182,
      "grad_norm": 0.76171875,
      "learning_rate": 5.73310193148199e-06,
      "loss": 0.5672,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4973
    },
    {
      "epoch": 0.6763666032091379,
      "grad_norm": 0.48828125,
      "learning_rate": 5.728808547915405e-06,
      "loss": 0.6768,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4974
    },
    {
      "epoch": 0.6765025836279576,
      "grad_norm": 0.40234375,
      "learning_rate": 5.724516127127598e-06,
      "loss": 0.7104,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4975
    },
    {
      "epoch": 0.6766385640467772,
      "grad_norm": 0.3671875,
      "learning_rate": 5.720224670086123e-06,
      "loss": 0.7147,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4976
    },
    {
      "epoch": 0.6767745444655969,
      "grad_norm": 0.330078125,
      "learning_rate": 5.715934177758333e-06,
      "loss": 0.6234,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4977
    },
    {
      "epoch": 0.6769105248844166,
      "grad_norm": 0.3984375,
      "learning_rate": 5.71164465111136e-06,
      "loss": 0.6725,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4978
    },
    {
      "epoch": 0.6770465053032363,
      "grad_norm": 0.3046875,
      "learning_rate": 5.707356091112108e-06,
      "loss": 0.4904,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4979
    },
    {
      "epoch": 0.677182485722056,
      "grad_norm": 0.35546875,
      "learning_rate": 5.70306849872728e-06,
      "loss": 0.6657,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4980
    },
    {
      "epoch": 0.6773184661408758,
      "grad_norm": 0.400390625,
      "learning_rate": 5.698781874923343e-06,
      "loss": 0.6348,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4981
    },
    {
      "epoch": 0.6774544465596954,
      "grad_norm": 0.60546875,
      "learning_rate": 5.694496220666563e-06,
      "loss": 0.6444,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4982
    },
    {
      "epoch": 0.6775904269785151,
      "grad_norm": 0.40234375,
      "learning_rate": 5.690211536922969e-06,
      "loss": 0.6125,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4983
    },
    {
      "epoch": 0.6777264073973348,
      "grad_norm": 0.322265625,
      "learning_rate": 5.685927824658391e-06,
      "loss": 0.5962,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4984
    },
    {
      "epoch": 0.6778623878161545,
      "grad_norm": 0.578125,
      "learning_rate": 5.681645084838422e-06,
      "loss": 0.6694,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4985
    },
    {
      "epoch": 0.6779983682349742,
      "grad_norm": 0.59375,
      "learning_rate": 5.677363318428451e-06,
      "loss": 0.4299,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4986
    },
    {
      "epoch": 0.6781343486537939,
      "grad_norm": 0.25,
      "learning_rate": 5.673082526393634e-06,
      "loss": 0.5016,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4987
    },
    {
      "epoch": 0.6782703290726135,
      "grad_norm": 0.4296875,
      "learning_rate": 5.668802709698919e-06,
      "loss": 0.6667,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4988
    },
    {
      "epoch": 0.6784063094914332,
      "grad_norm": 0.2373046875,
      "learning_rate": 5.664523869309023e-06,
      "loss": 0.5161,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4989
    },
    {
      "epoch": 0.6785422899102529,
      "grad_norm": 2.859375,
      "learning_rate": 5.660246006188458e-06,
      "loss": 0.8735,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4990
    },
    {
      "epoch": 0.6786782703290726,
      "grad_norm": 0.396484375,
      "learning_rate": 5.655969121301497e-06,
      "loss": 0.534,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4991
    },
    {
      "epoch": 0.6788142507478923,
      "grad_norm": 0.283203125,
      "learning_rate": 5.651693215612209e-06,
      "loss": 0.5768,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4992
    },
    {
      "epoch": 0.678950231166712,
      "grad_norm": 0.4765625,
      "learning_rate": 5.647418290084431e-06,
      "loss": 0.6632,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4993
    },
    {
      "epoch": 0.6790862115855317,
      "grad_norm": 0.3125,
      "learning_rate": 5.643144345681788e-06,
      "loss": 0.5312,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4994
    },
    {
      "epoch": 0.6792221920043514,
      "grad_norm": 0.390625,
      "learning_rate": 5.638871383367674e-06,
      "loss": 0.5418,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4995
    },
    {
      "epoch": 0.6793581724231711,
      "grad_norm": 0.578125,
      "learning_rate": 5.634599404105272e-06,
      "loss": 0.7209,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4996
    },
    {
      "epoch": 0.6794941528419908,
      "grad_norm": 0.283203125,
      "learning_rate": 5.6303284088575334e-06,
      "loss": 0.5552,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4997
    },
    {
      "epoch": 0.6796301332608105,
      "grad_norm": 0.2890625,
      "learning_rate": 5.6260583985872e-06,
      "loss": 0.4198,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4998
    },
    {
      "epoch": 0.6797661136796301,
      "grad_norm": 0.310546875,
      "learning_rate": 5.6217893742567745e-06,
      "loss": 0.5402,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 4999
    },
    {
      "epoch": 0.6799020940984498,
      "grad_norm": 0.38671875,
      "learning_rate": 5.617521336828556e-06,
      "loss": 0.7036,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5000
    },
    {
      "epoch": 0.6800380745172695,
      "grad_norm": 0.416015625,
      "learning_rate": 5.6132542872646055e-06,
      "loss": 0.7979,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5001
    },
    {
      "epoch": 0.6801740549360892,
      "grad_norm": 0.314453125,
      "learning_rate": 5.608988226526775e-06,
      "loss": 0.6007,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5002
    },
    {
      "epoch": 0.6803100353549089,
      "grad_norm": 0.58984375,
      "learning_rate": 5.604723155576679e-06,
      "loss": 0.8984,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5003
    },
    {
      "epoch": 0.6804460157737285,
      "grad_norm": 0.296875,
      "learning_rate": 5.600459075375726e-06,
      "loss": 0.4482,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5004
    },
    {
      "epoch": 0.6805819961925482,
      "grad_norm": 0.431640625,
      "learning_rate": 5.596195986885081e-06,
      "loss": 0.8146,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5005
    },
    {
      "epoch": 0.680717976611368,
      "grad_norm": 0.61328125,
      "learning_rate": 5.591933891065706e-06,
      "loss": 0.4639,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5006
    },
    {
      "epoch": 0.6808539570301877,
      "grad_norm": 0.5703125,
      "learning_rate": 5.587672788878322e-06,
      "loss": 0.5269,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5007
    },
    {
      "epoch": 0.6809899374490074,
      "grad_norm": 0.41015625,
      "learning_rate": 5.583412681283441e-06,
      "loss": 0.7795,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5008
    },
    {
      "epoch": 0.6811259178678271,
      "grad_norm": 1.5546875,
      "learning_rate": 5.579153569241335e-06,
      "loss": 0.8647,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5009
    },
    {
      "epoch": 0.6812618982866467,
      "grad_norm": 0.25390625,
      "learning_rate": 5.574895453712068e-06,
      "loss": 0.3966,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5010
    },
    {
      "epoch": 0.6813978787054664,
      "grad_norm": 0.24609375,
      "learning_rate": 5.570638335655462e-06,
      "loss": 0.3646,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5011
    },
    {
      "epoch": 0.6815338591242861,
      "grad_norm": 0.32421875,
      "learning_rate": 5.566382216031136e-06,
      "loss": 0.5368,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5012
    },
    {
      "epoch": 0.6816698395431058,
      "grad_norm": 0.36328125,
      "learning_rate": 5.5621270957984575e-06,
      "loss": 0.7231,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5013
    },
    {
      "epoch": 0.6818058199619255,
      "grad_norm": 0.283203125,
      "learning_rate": 5.5578729759165935e-06,
      "loss": 0.5565,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5014
    },
    {
      "epoch": 0.6819418003807451,
      "grad_norm": 0.27734375,
      "learning_rate": 5.553619857344468e-06,
      "loss": 0.5457,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5015
    },
    {
      "epoch": 0.6820777807995648,
      "grad_norm": 0.34375,
      "learning_rate": 5.549367741040792e-06,
      "loss": 0.5662,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5016
    },
    {
      "epoch": 0.6822137612183845,
      "grad_norm": 0.3515625,
      "learning_rate": 5.545116627964038e-06,
      "loss": 0.533,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5017
    },
    {
      "epoch": 0.6823497416372042,
      "grad_norm": 0.46875,
      "learning_rate": 5.540866519072463e-06,
      "loss": 0.5951,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5018
    },
    {
      "epoch": 0.682485722056024,
      "grad_norm": 0.40625,
      "learning_rate": 5.536617415324091e-06,
      "loss": 0.7493,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5019
    },
    {
      "epoch": 0.6826217024748437,
      "grad_norm": 0.2578125,
      "learning_rate": 5.532369317676726e-06,
      "loss": 0.4663,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5020
    },
    {
      "epoch": 0.6827576828936633,
      "grad_norm": 0.34375,
      "learning_rate": 5.528122227087935e-06,
      "loss": 0.5011,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5021
    },
    {
      "epoch": 0.682893663312483,
      "grad_norm": 0.58203125,
      "learning_rate": 5.52387614451507e-06,
      "loss": 0.7201,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5022
    },
    {
      "epoch": 0.6830296437313027,
      "grad_norm": 0.2216796875,
      "learning_rate": 5.519631070915245e-06,
      "loss": 0.5076,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5023
    },
    {
      "epoch": 0.6831656241501224,
      "grad_norm": 0.2138671875,
      "learning_rate": 5.515387007245357e-06,
      "loss": 0.3151,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5024
    },
    {
      "epoch": 0.6833016045689421,
      "grad_norm": 0.44140625,
      "learning_rate": 5.511143954462064e-06,
      "loss": 0.8215,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5025
    },
    {
      "epoch": 0.6834375849877617,
      "grad_norm": 0.703125,
      "learning_rate": 5.506901913521808e-06,
      "loss": 0.6452,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5026
    },
    {
      "epoch": 0.6835735654065814,
      "grad_norm": 0.43359375,
      "learning_rate": 5.502660885380791e-06,
      "loss": 0.5768,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5027
    },
    {
      "epoch": 0.6837095458254011,
      "grad_norm": 0.328125,
      "learning_rate": 5.498420870994997e-06,
      "loss": 0.7179,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5028
    },
    {
      "epoch": 0.6838455262442208,
      "grad_norm": 0.58984375,
      "learning_rate": 5.494181871320172e-06,
      "loss": 0.7562,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5029
    },
    {
      "epoch": 0.6839815066630405,
      "grad_norm": 0.431640625,
      "learning_rate": 5.489943887311846e-06,
      "loss": 0.7123,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5030
    },
    {
      "epoch": 0.6841174870818603,
      "grad_norm": 0.361328125,
      "learning_rate": 5.485706919925304e-06,
      "loss": 0.4725,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5031
    },
    {
      "epoch": 0.68425346750068,
      "grad_norm": 0.279296875,
      "learning_rate": 5.481470970115619e-06,
      "loss": 0.5244,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5032
    },
    {
      "epoch": 0.6843894479194996,
      "grad_norm": 0.640625,
      "learning_rate": 5.477236038837617e-06,
      "loss": 0.761,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5033
    },
    {
      "epoch": 0.6845254283383193,
      "grad_norm": 0.31640625,
      "learning_rate": 5.473002127045912e-06,
      "loss": 0.6497,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5034
    },
    {
      "epoch": 0.684661408757139,
      "grad_norm": 0.28515625,
      "learning_rate": 5.468769235694871e-06,
      "loss": 0.4712,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5035
    },
    {
      "epoch": 0.6847973891759587,
      "grad_norm": 0.40625,
      "learning_rate": 5.464537365738646e-06,
      "loss": 0.6014,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5036
    },
    {
      "epoch": 0.6849333695947784,
      "grad_norm": 0.4921875,
      "learning_rate": 5.460306518131146e-06,
      "loss": 0.7171,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5037
    },
    {
      "epoch": 0.685069350013598,
      "grad_norm": 0.294921875,
      "learning_rate": 5.45607669382606e-06,
      "loss": 0.5373,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5038
    },
    {
      "epoch": 0.6852053304324177,
      "grad_norm": 0.43359375,
      "learning_rate": 5.451847893776845e-06,
      "loss": 0.7936,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5039
    },
    {
      "epoch": 0.6853413108512374,
      "grad_norm": 0.33984375,
      "learning_rate": 5.447620118936718e-06,
      "loss": 0.7248,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5040
    },
    {
      "epoch": 0.6854772912700571,
      "grad_norm": 0.3671875,
      "learning_rate": 5.443393370258677e-06,
      "loss": 0.6062,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5041
    },
    {
      "epoch": 0.6856132716888768,
      "grad_norm": 0.4921875,
      "learning_rate": 5.439167648695476e-06,
      "loss": 0.8309,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5042
    },
    {
      "epoch": 0.6857492521076964,
      "grad_norm": 0.357421875,
      "learning_rate": 5.434942955199653e-06,
      "loss": 0.6497,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5043
    },
    {
      "epoch": 0.6858852325265162,
      "grad_norm": 0.427734375,
      "learning_rate": 5.430719290723495e-06,
      "loss": 0.8587,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5044
    },
    {
      "epoch": 0.6860212129453359,
      "grad_norm": 0.67578125,
      "learning_rate": 5.426496656219078e-06,
      "loss": 0.6419,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5045
    },
    {
      "epoch": 0.6861571933641556,
      "grad_norm": 1.453125,
      "learning_rate": 5.422275052638227e-06,
      "loss": 0.6668,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5046
    },
    {
      "epoch": 0.6862931737829753,
      "grad_norm": 0.50390625,
      "learning_rate": 5.418054480932551e-06,
      "loss": 0.8457,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5047
    },
    {
      "epoch": 0.686429154201795,
      "grad_norm": 0.228515625,
      "learning_rate": 5.4138349420534085e-06,
      "loss": 0.4325,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5048
    },
    {
      "epoch": 0.6865651346206146,
      "grad_norm": 0.4921875,
      "learning_rate": 5.409616436951948e-06,
      "loss": 0.714,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5049
    },
    {
      "epoch": 0.6867011150394343,
      "grad_norm": 0.37890625,
      "learning_rate": 5.405398966579059e-06,
      "loss": 0.5519,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5050
    },
    {
      "epoch": 0.686837095458254,
      "grad_norm": 0.337890625,
      "learning_rate": 5.401182531885421e-06,
      "loss": 0.6621,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5051
    },
    {
      "epoch": 0.6869730758770737,
      "grad_norm": 0.65234375,
      "learning_rate": 5.396967133821461e-06,
      "loss": 0.6069,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5052
    },
    {
      "epoch": 0.6871090562958934,
      "grad_norm": 0.26953125,
      "learning_rate": 5.392752773337391e-06,
      "loss": 0.453,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5053
    },
    {
      "epoch": 0.687245036714713,
      "grad_norm": 0.4765625,
      "learning_rate": 5.38853945138317e-06,
      "loss": 0.3957,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5054
    },
    {
      "epoch": 0.6873810171335327,
      "grad_norm": 0.41015625,
      "learning_rate": 5.384327168908541e-06,
      "loss": 0.6589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5055
    },
    {
      "epoch": 0.6875169975523525,
      "grad_norm": 0.53125,
      "learning_rate": 5.380115926862997e-06,
      "loss": 0.5133,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5056
    },
    {
      "epoch": 0.6876529779711722,
      "grad_norm": 0.376953125,
      "learning_rate": 5.375905726195809e-06,
      "loss": 0.6024,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5057
    },
    {
      "epoch": 0.6877889583899919,
      "grad_norm": 0.326171875,
      "learning_rate": 5.371696567855999e-06,
      "loss": 0.7031,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5058
    },
    {
      "epoch": 0.6879249388088116,
      "grad_norm": 0.361328125,
      "learning_rate": 5.3674884527923755e-06,
      "loss": 0.5732,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5059
    },
    {
      "epoch": 0.6880609192276312,
      "grad_norm": 0.314453125,
      "learning_rate": 5.3632813819534864e-06,
      "loss": 0.5924,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5060
    },
    {
      "epoch": 0.6881968996464509,
      "grad_norm": 0.357421875,
      "learning_rate": 5.3590753562876685e-06,
      "loss": 0.624,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5061
    },
    {
      "epoch": 0.6883328800652706,
      "grad_norm": 0.369140625,
      "learning_rate": 5.354870376743e-06,
      "loss": 0.6702,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5062
    },
    {
      "epoch": 0.6884688604840903,
      "grad_norm": 0.359375,
      "learning_rate": 5.350666444267345e-06,
      "loss": 0.6434,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5063
    },
    {
      "epoch": 0.68860484090291,
      "grad_norm": 0.341796875,
      "learning_rate": 5.3464635598083126e-06,
      "loss": 0.7109,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5064
    },
    {
      "epoch": 0.6887408213217296,
      "grad_norm": 0.4375,
      "learning_rate": 5.342261724313292e-06,
      "loss": 0.6792,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5065
    },
    {
      "epoch": 0.6888768017405493,
      "grad_norm": 0.4375,
      "learning_rate": 5.338060938729422e-06,
      "loss": 0.7642,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5066
    },
    {
      "epoch": 0.689012782159369,
      "grad_norm": 0.236328125,
      "learning_rate": 5.333861204003616e-06,
      "loss": 0.4508,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5067
    },
    {
      "epoch": 0.6891487625781887,
      "grad_norm": 0.369140625,
      "learning_rate": 5.329662521082539e-06,
      "loss": 0.6659,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5068
    },
    {
      "epoch": 0.6892847429970085,
      "grad_norm": 0.41015625,
      "learning_rate": 5.325464890912635e-06,
      "loss": 0.6585,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5069
    },
    {
      "epoch": 0.6894207234158282,
      "grad_norm": 0.46484375,
      "learning_rate": 5.3212683144400915e-06,
      "loss": 0.8454,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5070
    },
    {
      "epoch": 0.6895567038346478,
      "grad_norm": 0.3515625,
      "learning_rate": 5.317072792610876e-06,
      "loss": 0.5775,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5071
    },
    {
      "epoch": 0.6896926842534675,
      "grad_norm": 0.353515625,
      "learning_rate": 5.312878326370703e-06,
      "loss": 0.7703,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5072
    },
    {
      "epoch": 0.6898286646722872,
      "grad_norm": 0.65234375,
      "learning_rate": 5.308684916665064e-06,
      "loss": 0.6899,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5073
    },
    {
      "epoch": 0.6899646450911069,
      "grad_norm": 0.322265625,
      "learning_rate": 5.304492564439198e-06,
      "loss": 0.4971,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5074
    },
    {
      "epoch": 0.6901006255099266,
      "grad_norm": 0.376953125,
      "learning_rate": 5.3003012706381195e-06,
      "loss": 0.6817,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5075
    },
    {
      "epoch": 0.6902366059287462,
      "grad_norm": 0.30078125,
      "learning_rate": 5.296111036206589e-06,
      "loss": 0.6484,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5076
    },
    {
      "epoch": 0.6903725863475659,
      "grad_norm": 0.412109375,
      "learning_rate": 5.291921862089144e-06,
      "loss": 0.6077,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5077
    },
    {
      "epoch": 0.6905085667663856,
      "grad_norm": 0.51953125,
      "learning_rate": 5.287733749230071e-06,
      "loss": 0.625,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5078
    },
    {
      "epoch": 0.6906445471852053,
      "grad_norm": 0.30859375,
      "learning_rate": 5.283546698573425e-06,
      "loss": 0.5571,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5079
    },
    {
      "epoch": 0.690780527604025,
      "grad_norm": 0.34375,
      "learning_rate": 5.279360711063015e-06,
      "loss": 0.5667,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5080
    },
    {
      "epoch": 0.6909165080228447,
      "grad_norm": 0.328125,
      "learning_rate": 5.275175787642419e-06,
      "loss": 0.624,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5081
    },
    {
      "epoch": 0.6910524884416644,
      "grad_norm": 0.376953125,
      "learning_rate": 5.270991929254963e-06,
      "loss": 0.5705,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5082
    },
    {
      "epoch": 0.6911884688604841,
      "grad_norm": 0.255859375,
      "learning_rate": 5.266809136843747e-06,
      "loss": 0.4517,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5083
    },
    {
      "epoch": 0.6913244492793038,
      "grad_norm": 0.55859375,
      "learning_rate": 5.262627411351615e-06,
      "loss": 0.793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5084
    },
    {
      "epoch": 0.6914604296981235,
      "grad_norm": 0.431640625,
      "learning_rate": 5.258446753721189e-06,
      "loss": 0.6099,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5085
    },
    {
      "epoch": 0.6915964101169432,
      "grad_norm": 0.796875,
      "learning_rate": 5.254267164894831e-06,
      "loss": 0.5954,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5086
    },
    {
      "epoch": 0.6917323905357629,
      "grad_norm": 0.298828125,
      "learning_rate": 5.250088645814679e-06,
      "loss": 0.6553,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5087
    },
    {
      "epoch": 0.6918683709545825,
      "grad_norm": 0.36328125,
      "learning_rate": 5.245911197422616e-06,
      "loss": 0.5728,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5088
    },
    {
      "epoch": 0.6920043513734022,
      "grad_norm": 0.43359375,
      "learning_rate": 5.241734820660296e-06,
      "loss": 0.6689,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5089
    },
    {
      "epoch": 0.6921403317922219,
      "grad_norm": 0.61328125,
      "learning_rate": 5.23755951646912e-06,
      "loss": 0.4045,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5090
    },
    {
      "epoch": 0.6922763122110416,
      "grad_norm": 0.39453125,
      "learning_rate": 5.233385285790258e-06,
      "loss": 0.6198,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5091
    },
    {
      "epoch": 0.6924122926298613,
      "grad_norm": 0.3203125,
      "learning_rate": 5.229212129564625e-06,
      "loss": 0.6102,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5092
    },
    {
      "epoch": 0.6925482730486809,
      "grad_norm": 0.326171875,
      "learning_rate": 5.225040048732911e-06,
      "loss": 0.6859,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5093
    },
    {
      "epoch": 0.6926842534675007,
      "grad_norm": 0.1962890625,
      "learning_rate": 5.220869044235545e-06,
      "loss": 0.2832,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5094
    },
    {
      "epoch": 0.6928202338863204,
      "grad_norm": 0.390625,
      "learning_rate": 5.216699117012729e-06,
      "loss": 0.771,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5095
    },
    {
      "epoch": 0.6929562143051401,
      "grad_norm": 0.63671875,
      "learning_rate": 5.212530268004411e-06,
      "loss": 0.7248,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5096
    },
    {
      "epoch": 0.6930921947239598,
      "grad_norm": 0.69140625,
      "learning_rate": 5.208362498150305e-06,
      "loss": 0.6506,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5097
    },
    {
      "epoch": 0.6932281751427795,
      "grad_norm": 0.515625,
      "learning_rate": 5.2041958083898724e-06,
      "loss": 0.8501,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5098
    },
    {
      "epoch": 0.6933641555615991,
      "grad_norm": 0.30078125,
      "learning_rate": 5.200030199662338e-06,
      "loss": 0.5658,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5099
    },
    {
      "epoch": 0.6935001359804188,
      "grad_norm": 0.33984375,
      "learning_rate": 5.195865672906685e-06,
      "loss": 0.522,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5100
    },
    {
      "epoch": 0.6936361163992385,
      "grad_norm": 0.396484375,
      "learning_rate": 5.191702229061641e-06,
      "loss": 0.7595,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5101
    },
    {
      "epoch": 0.6937720968180582,
      "grad_norm": 0.28125,
      "learning_rate": 5.187539869065706e-06,
      "loss": 0.5093,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5102
    },
    {
      "epoch": 0.6939080772368779,
      "grad_norm": 0.7734375,
      "learning_rate": 5.183378593857117e-06,
      "loss": 0.7145,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5103
    },
    {
      "epoch": 0.6940440576556975,
      "grad_norm": 0.3203125,
      "learning_rate": 5.179218404373886e-06,
      "loss": 0.7018,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5104
    },
    {
      "epoch": 0.6941800380745172,
      "grad_norm": 0.7109375,
      "learning_rate": 5.175059301553762e-06,
      "loss": 0.7549,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5105
    },
    {
      "epoch": 0.6943160184933369,
      "grad_norm": 0.28515625,
      "learning_rate": 5.170901286334264e-06,
      "loss": 0.4492,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5106
    },
    {
      "epoch": 0.6944519989121567,
      "grad_norm": 0.3984375,
      "learning_rate": 5.1667443596526555e-06,
      "loss": 0.5754,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5107
    },
    {
      "epoch": 0.6945879793309764,
      "grad_norm": 0.375,
      "learning_rate": 5.162588522445961e-06,
      "loss": 0.5999,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5108
    },
    {
      "epoch": 0.6947239597497961,
      "grad_norm": 0.353515625,
      "learning_rate": 5.158433775650955e-06,
      "loss": 0.695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5109
    },
    {
      "epoch": 0.6948599401686157,
      "grad_norm": 0.6953125,
      "learning_rate": 5.154280120204173e-06,
      "loss": 0.5628,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5110
    },
    {
      "epoch": 0.6949959205874354,
      "grad_norm": 0.37109375,
      "learning_rate": 5.150127557041891e-06,
      "loss": 0.5814,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5111
    },
    {
      "epoch": 0.6951319010062551,
      "grad_norm": 0.5078125,
      "learning_rate": 5.145976087100157e-06,
      "loss": 0.652,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5112
    },
    {
      "epoch": 0.6952678814250748,
      "grad_norm": 0.341796875,
      "learning_rate": 5.141825711314754e-06,
      "loss": 0.6099,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5113
    },
    {
      "epoch": 0.6954038618438945,
      "grad_norm": 0.451171875,
      "learning_rate": 5.137676430621237e-06,
      "loss": 0.7539,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5114
    },
    {
      "epoch": 0.6955398422627141,
      "grad_norm": 0.72265625,
      "learning_rate": 5.133528245954895e-06,
      "loss": 0.8115,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5115
    },
    {
      "epoch": 0.6956758226815338,
      "grad_norm": 0.369140625,
      "learning_rate": 5.129381158250789e-06,
      "loss": 0.5075,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5116
    },
    {
      "epoch": 0.6958118031003535,
      "grad_norm": 0.42578125,
      "learning_rate": 5.125235168443714e-06,
      "loss": 0.7528,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5117
    },
    {
      "epoch": 0.6959477835191732,
      "grad_norm": 0.388671875,
      "learning_rate": 5.121090277468235e-06,
      "loss": 0.7018,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5118
    },
    {
      "epoch": 0.696083763937993,
      "grad_norm": 0.439453125,
      "learning_rate": 5.116946486258654e-06,
      "loss": 0.4635,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5119
    },
    {
      "epoch": 0.6962197443568127,
      "grad_norm": 0.44140625,
      "learning_rate": 5.11280379574904e-06,
      "loss": 0.7318,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5120
    },
    {
      "epoch": 0.6963557247756323,
      "grad_norm": 0.443359375,
      "learning_rate": 5.108662206873196e-06,
      "loss": 0.6099,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5121
    },
    {
      "epoch": 0.696491705194452,
      "grad_norm": 0.259765625,
      "learning_rate": 5.104521720564698e-06,
      "loss": 0.4969,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5122
    },
    {
      "epoch": 0.6966276856132717,
      "grad_norm": 0.216796875,
      "learning_rate": 5.100382337756853e-06,
      "loss": 0.4049,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5123
    },
    {
      "epoch": 0.6967636660320914,
      "grad_norm": 0.349609375,
      "learning_rate": 5.0962440593827355e-06,
      "loss": 0.5866,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5124
    },
    {
      "epoch": 0.6968996464509111,
      "grad_norm": 0.30859375,
      "learning_rate": 5.0921068863751565e-06,
      "loss": 0.4772,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5125
    },
    {
      "epoch": 0.6970356268697308,
      "grad_norm": 0.58203125,
      "learning_rate": 5.087970819666693e-06,
      "loss": 0.5645,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5126
    },
    {
      "epoch": 0.6971716072885504,
      "grad_norm": 0.37890625,
      "learning_rate": 5.083835860189659e-06,
      "loss": 0.6305,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5127
    },
    {
      "epoch": 0.6973075877073701,
      "grad_norm": 0.333984375,
      "learning_rate": 5.0797020088761306e-06,
      "loss": 0.562,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5128
    },
    {
      "epoch": 0.6974435681261898,
      "grad_norm": 0.302734375,
      "learning_rate": 5.075569266657921e-06,
      "loss": 0.6147,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5129
    },
    {
      "epoch": 0.6975795485450095,
      "grad_norm": 0.33203125,
      "learning_rate": 5.0714376344666095e-06,
      "loss": 0.5589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5130
    },
    {
      "epoch": 0.6977155289638292,
      "grad_norm": 0.439453125,
      "learning_rate": 5.06730711323351e-06,
      "loss": 0.6201,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5131
    },
    {
      "epoch": 0.697851509382649,
      "grad_norm": 0.427734375,
      "learning_rate": 5.0631777038896965e-06,
      "loss": 0.6859,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5132
    },
    {
      "epoch": 0.6979874898014686,
      "grad_norm": 0.6953125,
      "learning_rate": 5.059049407365984e-06,
      "loss": 0.6797,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5133
    },
    {
      "epoch": 0.6981234702202883,
      "grad_norm": 0.46484375,
      "learning_rate": 5.054922224592947e-06,
      "loss": 0.6143,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5134
    },
    {
      "epoch": 0.698259450639108,
      "grad_norm": 0.419921875,
      "learning_rate": 5.050796156500895e-06,
      "loss": 0.6375,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5135
    },
    {
      "epoch": 0.6983954310579277,
      "grad_norm": 0.66796875,
      "learning_rate": 5.0466712040199015e-06,
      "loss": 0.7388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5136
    },
    {
      "epoch": 0.6985314114767474,
      "grad_norm": 0.6328125,
      "learning_rate": 5.042547368079776e-06,
      "loss": 0.4928,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5137
    },
    {
      "epoch": 0.698667391895567,
      "grad_norm": 0.451171875,
      "learning_rate": 5.038424649610087e-06,
      "loss": 0.9425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5138
    },
    {
      "epoch": 0.6988033723143867,
      "grad_norm": 0.74609375,
      "learning_rate": 5.03430304954014e-06,
      "loss": 0.7562,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5139
    },
    {
      "epoch": 0.6989393527332064,
      "grad_norm": 0.54296875,
      "learning_rate": 5.030182568799e-06,
      "loss": 0.5664,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5140
    },
    {
      "epoch": 0.6990753331520261,
      "grad_norm": 0.37890625,
      "learning_rate": 5.026063208315467e-06,
      "loss": 0.6245,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5141
    },
    {
      "epoch": 0.6992113135708458,
      "grad_norm": 0.41015625,
      "learning_rate": 5.021944969018102e-06,
      "loss": 0.777,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5142
    },
    {
      "epoch": 0.6993472939896654,
      "grad_norm": 0.48046875,
      "learning_rate": 5.017827851835199e-06,
      "loss": 0.8659,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5143
    },
    {
      "epoch": 0.6994832744084851,
      "grad_norm": 0.33984375,
      "learning_rate": 5.0137118576948144e-06,
      "loss": 0.5138,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5144
    },
    {
      "epoch": 0.6996192548273049,
      "grad_norm": 0.31640625,
      "learning_rate": 5.0095969875247365e-06,
      "loss": 0.5037,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5145
    },
    {
      "epoch": 0.6997552352461246,
      "grad_norm": 0.69921875,
      "learning_rate": 5.005483242252514e-06,
      "loss": 0.7915,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5146
    },
    {
      "epoch": 0.6998912156649443,
      "grad_norm": 0.384765625,
      "learning_rate": 5.001370622805428e-06,
      "loss": 0.5338,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5147
    },
    {
      "epoch": 0.700027196083764,
      "grad_norm": 0.330078125,
      "learning_rate": 4.997259130110522e-06,
      "loss": 0.5308,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5148
    },
    {
      "epoch": 0.7001631765025836,
      "grad_norm": 0.2734375,
      "learning_rate": 4.993148765094567e-06,
      "loss": 0.4515,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5149
    },
    {
      "epoch": 0.7002991569214033,
      "grad_norm": 0.52734375,
      "learning_rate": 4.9890395286841e-06,
      "loss": 0.3709,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5150
    },
    {
      "epoch": 0.700435137340223,
      "grad_norm": 0.58984375,
      "learning_rate": 4.9849314218053825e-06,
      "loss": 0.7752,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5151
    },
    {
      "epoch": 0.7005711177590427,
      "grad_norm": 0.3125,
      "learning_rate": 4.9808244453844414e-06,
      "loss": 0.6435,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5152
    },
    {
      "epoch": 0.7007070981778624,
      "grad_norm": 0.30078125,
      "learning_rate": 4.976718600347031e-06,
      "loss": 0.5239,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5153
    },
    {
      "epoch": 0.700843078596682,
      "grad_norm": 0.435546875,
      "learning_rate": 4.972613887618667e-06,
      "loss": 0.5441,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5154
    },
    {
      "epoch": 0.7009790590155017,
      "grad_norm": 0.44921875,
      "learning_rate": 4.968510308124595e-06,
      "loss": 0.5904,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5155
    },
    {
      "epoch": 0.7011150394343214,
      "grad_norm": 0.419921875,
      "learning_rate": 4.964407862789817e-06,
      "loss": 0.786,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5156
    },
    {
      "epoch": 0.7012510198531412,
      "grad_norm": 0.5,
      "learning_rate": 4.96030655253907e-06,
      "loss": 0.755,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5157
    },
    {
      "epoch": 0.7013870002719609,
      "grad_norm": 0.5,
      "learning_rate": 4.95620637829684e-06,
      "loss": 0.7388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5158
    },
    {
      "epoch": 0.7015229806907806,
      "grad_norm": 0.3203125,
      "learning_rate": 4.952107340987365e-06,
      "loss": 0.7178,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5159
    },
    {
      "epoch": 0.7016589611096002,
      "grad_norm": 1.296875,
      "learning_rate": 4.9480094415346046e-06,
      "loss": 0.7847,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5160
    },
    {
      "epoch": 0.7017949415284199,
      "grad_norm": 0.390625,
      "learning_rate": 4.943912680862288e-06,
      "loss": 0.603,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5161
    },
    {
      "epoch": 0.7019309219472396,
      "grad_norm": 0.69140625,
      "learning_rate": 4.939817059893864e-06,
      "loss": 0.9312,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5162
    },
    {
      "epoch": 0.7020669023660593,
      "grad_norm": 0.55078125,
      "learning_rate": 4.935722579552546e-06,
      "loss": 0.4702,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5163
    },
    {
      "epoch": 0.702202882784879,
      "grad_norm": 0.45703125,
      "learning_rate": 4.93162924076127e-06,
      "loss": 0.6466,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5164
    },
    {
      "epoch": 0.7023388632036986,
      "grad_norm": 0.287109375,
      "learning_rate": 4.927537044442733e-06,
      "loss": 0.6139,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5165
    },
    {
      "epoch": 0.7024748436225183,
      "grad_norm": 0.3828125,
      "learning_rate": 4.923445991519359e-06,
      "loss": 0.6888,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5166
    },
    {
      "epoch": 0.702610824041338,
      "grad_norm": 0.625,
      "learning_rate": 4.919356082913329e-06,
      "loss": 0.8398,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5167
    },
    {
      "epoch": 0.7027468044601577,
      "grad_norm": 0.349609375,
      "learning_rate": 4.9152673195465505e-06,
      "loss": 0.6501,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5168
    },
    {
      "epoch": 0.7028827848789774,
      "grad_norm": 0.3671875,
      "learning_rate": 4.9111797023406884e-06,
      "loss": 0.5659,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5169
    },
    {
      "epoch": 0.7030187652977972,
      "grad_norm": 0.17578125,
      "learning_rate": 4.907093232217133e-06,
      "loss": 0.2573,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5170
    },
    {
      "epoch": 0.7031547457166168,
      "grad_norm": 0.40234375,
      "learning_rate": 4.903007910097034e-06,
      "loss": 0.7516,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5171
    },
    {
      "epoch": 0.7032907261354365,
      "grad_norm": 0.322265625,
      "learning_rate": 4.898923736901265e-06,
      "loss": 0.422,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5172
    },
    {
      "epoch": 0.7034267065542562,
      "grad_norm": 0.447265625,
      "learning_rate": 4.894840713550454e-06,
      "loss": 0.5671,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5173
    },
    {
      "epoch": 0.7035626869730759,
      "grad_norm": 0.3984375,
      "learning_rate": 4.89075884096496e-06,
      "loss": 0.4658,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5174
    },
    {
      "epoch": 0.7036986673918956,
      "grad_norm": 0.37109375,
      "learning_rate": 4.886678120064892e-06,
      "loss": 0.7269,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5175
    },
    {
      "epoch": 0.7038346478107153,
      "grad_norm": 0.41796875,
      "learning_rate": 4.882598551770088e-06,
      "loss": 0.6862,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5176
    },
    {
      "epoch": 0.7039706282295349,
      "grad_norm": 0.3203125,
      "learning_rate": 4.87852013700014e-06,
      "loss": 0.6364,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5177
    },
    {
      "epoch": 0.7041066086483546,
      "grad_norm": 0.41796875,
      "learning_rate": 4.874442876674366e-06,
      "loss": 0.641,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5178
    },
    {
      "epoch": 0.7042425890671743,
      "grad_norm": 0.36328125,
      "learning_rate": 4.870366771711835e-06,
      "loss": 0.5745,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5179
    },
    {
      "epoch": 0.704378569485994,
      "grad_norm": 0.302734375,
      "learning_rate": 4.866291823031346e-06,
      "loss": 0.5724,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5180
    },
    {
      "epoch": 0.7045145499048137,
      "grad_norm": 0.4140625,
      "learning_rate": 4.86221803155145e-06,
      "loss": 0.5892,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5181
    },
    {
      "epoch": 0.7046505303236335,
      "grad_norm": 0.404296875,
      "learning_rate": 4.8581453981904205e-06,
      "loss": 0.6937,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5182
    },
    {
      "epoch": 0.7047865107424531,
      "grad_norm": 0.41796875,
      "learning_rate": 4.854073923866286e-06,
      "loss": 0.7376,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5183
    },
    {
      "epoch": 0.7049224911612728,
      "grad_norm": 0.310546875,
      "learning_rate": 4.850003609496799e-06,
      "loss": 0.556,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5184
    },
    {
      "epoch": 0.7050584715800925,
      "grad_norm": 0.275390625,
      "learning_rate": 4.845934455999468e-06,
      "loss": 0.4772,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5185
    },
    {
      "epoch": 0.7051944519989122,
      "grad_norm": 0.359375,
      "learning_rate": 4.8418664642915204e-06,
      "loss": 0.5563,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5186
    },
    {
      "epoch": 0.7053304324177319,
      "grad_norm": 0.466796875,
      "learning_rate": 4.8377996352899395e-06,
      "loss": 0.7257,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5187
    },
    {
      "epoch": 0.7054664128365515,
      "grad_norm": 0.306640625,
      "learning_rate": 4.83373396991143e-06,
      "loss": 0.7137,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5188
    },
    {
      "epoch": 0.7056023932553712,
      "grad_norm": 0.412109375,
      "learning_rate": 4.829669469072451e-06,
      "loss": 0.6331,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5189
    },
    {
      "epoch": 0.7057383736741909,
      "grad_norm": 0.404296875,
      "learning_rate": 4.8256061336891826e-06,
      "loss": 0.7585,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5190
    },
    {
      "epoch": 0.7058743540930106,
      "grad_norm": 0.6328125,
      "learning_rate": 4.821543964677559e-06,
      "loss": 0.756,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5191
    },
    {
      "epoch": 0.7060103345118303,
      "grad_norm": 0.3984375,
      "learning_rate": 4.817482962953233e-06,
      "loss": 0.6143,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5192
    },
    {
      "epoch": 0.7061463149306499,
      "grad_norm": 0.421875,
      "learning_rate": 4.813423129431612e-06,
      "loss": 0.5776,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5193
    },
    {
      "epoch": 0.7062822953494696,
      "grad_norm": 0.2216796875,
      "learning_rate": 4.809364465027827e-06,
      "loss": 0.4152,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5194
    },
    {
      "epoch": 0.7064182757682894,
      "grad_norm": 0.3125,
      "learning_rate": 4.8053069706567555e-06,
      "loss": 0.5508,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5195
    },
    {
      "epoch": 0.7065542561871091,
      "grad_norm": 0.208984375,
      "learning_rate": 4.801250647233001e-06,
      "loss": 0.4479,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5196
    },
    {
      "epoch": 0.7066902366059288,
      "grad_norm": 0.70703125,
      "learning_rate": 4.7971954956709135e-06,
      "loss": 0.5257,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5197
    },
    {
      "epoch": 0.7068262170247485,
      "grad_norm": 0.7890625,
      "learning_rate": 4.793141516884568e-06,
      "loss": 0.6929,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5198
    },
    {
      "epoch": 0.7069621974435681,
      "grad_norm": 0.5390625,
      "learning_rate": 4.789088711787787e-06,
      "loss": 0.8446,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5199
    },
    {
      "epoch": 0.7070981778623878,
      "grad_norm": 0.66015625,
      "learning_rate": 4.785037081294117e-06,
      "loss": 0.569,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5200
    },
    {
      "epoch": 0.7072341582812075,
      "grad_norm": 0.380859375,
      "learning_rate": 4.7809866263168505e-06,
      "loss": 0.6652,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5201
    },
    {
      "epoch": 0.7073701387000272,
      "grad_norm": 0.58984375,
      "learning_rate": 4.776937347769003e-06,
      "loss": 0.5265,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5202
    },
    {
      "epoch": 0.7075061191188469,
      "grad_norm": 0.349609375,
      "learning_rate": 4.772889246563338e-06,
      "loss": 0.6831,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5203
    },
    {
      "epoch": 0.7076420995376665,
      "grad_norm": 0.2890625,
      "learning_rate": 4.768842323612341e-06,
      "loss": 0.3706,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5204
    },
    {
      "epoch": 0.7077780799564862,
      "grad_norm": 0.58203125,
      "learning_rate": 4.764796579828243e-06,
      "loss": 0.527,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5205
    },
    {
      "epoch": 0.7079140603753059,
      "grad_norm": 0.2490234375,
      "learning_rate": 4.760752016122999e-06,
      "loss": 0.4281,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5206
    },
    {
      "epoch": 0.7080500407941257,
      "grad_norm": 0.359375,
      "learning_rate": 4.7567086334083105e-06,
      "loss": 0.598,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5207
    },
    {
      "epoch": 0.7081860212129454,
      "grad_norm": 0.58984375,
      "learning_rate": 4.752666432595596e-06,
      "loss": 0.6916,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5208
    },
    {
      "epoch": 0.7083220016317651,
      "grad_norm": 0.31640625,
      "learning_rate": 4.748625414596026e-06,
      "loss": 0.5736,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5209
    },
    {
      "epoch": 0.7084579820505847,
      "grad_norm": 0.5,
      "learning_rate": 4.7445855803204885e-06,
      "loss": 0.6287,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5210
    },
    {
      "epoch": 0.7085939624694044,
      "grad_norm": 0.41015625,
      "learning_rate": 4.740546930679617e-06,
      "loss": 0.4574,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5211
    },
    {
      "epoch": 0.7087299428882241,
      "grad_norm": 0.2431640625,
      "learning_rate": 4.736509466583767e-06,
      "loss": 0.5228,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5212
    },
    {
      "epoch": 0.7088659233070438,
      "grad_norm": 0.3203125,
      "learning_rate": 4.732473188943039e-06,
      "loss": 0.5985,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5213
    },
    {
      "epoch": 0.7090019037258635,
      "grad_norm": 1.2890625,
      "learning_rate": 4.7284380986672505e-06,
      "loss": 0.9307,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5214
    },
    {
      "epoch": 0.7091378841446832,
      "grad_norm": 0.3515625,
      "learning_rate": 4.724404196665969e-06,
      "loss": 0.7979,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5215
    },
    {
      "epoch": 0.7092738645635028,
      "grad_norm": 0.5234375,
      "learning_rate": 4.720371483848482e-06,
      "loss": 0.6182,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5216
    },
    {
      "epoch": 0.7094098449823225,
      "grad_norm": 0.6875,
      "learning_rate": 4.716339961123807e-06,
      "loss": 0.9754,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5217
    },
    {
      "epoch": 0.7095458254011422,
      "grad_norm": 0.392578125,
      "learning_rate": 4.712309629400704e-06,
      "loss": 0.7925,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5218
    },
    {
      "epoch": 0.7096818058199619,
      "grad_norm": 0.62890625,
      "learning_rate": 4.708280489587659e-06,
      "loss": 0.557,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5219
    },
    {
      "epoch": 0.7098177862387817,
      "grad_norm": 0.419921875,
      "learning_rate": 4.704252542592891e-06,
      "loss": 0.8234,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5220
    },
    {
      "epoch": 0.7099537666576013,
      "grad_norm": 0.62890625,
      "learning_rate": 4.700225789324343e-06,
      "loss": 0.5928,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5221
    },
    {
      "epoch": 0.710089747076421,
      "grad_norm": 0.2578125,
      "learning_rate": 4.696200230689701e-06,
      "loss": 0.3905,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5222
    },
    {
      "epoch": 0.7102257274952407,
      "grad_norm": 0.404296875,
      "learning_rate": 4.692175867596368e-06,
      "loss": 0.8926,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5223
    },
    {
      "epoch": 0.7103617079140604,
      "grad_norm": 0.33203125,
      "learning_rate": 4.688152700951493e-06,
      "loss": 0.5231,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5224
    },
    {
      "epoch": 0.7104976883328801,
      "grad_norm": 0.390625,
      "learning_rate": 4.684130731661938e-06,
      "loss": 0.6123,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5225
    },
    {
      "epoch": 0.7106336687516998,
      "grad_norm": 0.53515625,
      "learning_rate": 4.680109960634312e-06,
      "loss": 0.6784,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5226
    },
    {
      "epoch": 0.7107696491705194,
      "grad_norm": 0.408203125,
      "learning_rate": 4.676090388774937e-06,
      "loss": 0.7253,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5227
    },
    {
      "epoch": 0.7109056295893391,
      "grad_norm": 0.5546875,
      "learning_rate": 4.672072016989885e-06,
      "loss": 0.4137,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5228
    },
    {
      "epoch": 0.7110416100081588,
      "grad_norm": 0.2890625,
      "learning_rate": 4.668054846184935e-06,
      "loss": 0.4157,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5229
    },
    {
      "epoch": 0.7111775904269785,
      "grad_norm": 0.48046875,
      "learning_rate": 4.664038877265615e-06,
      "loss": 0.7458,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5230
    },
    {
      "epoch": 0.7113135708457982,
      "grad_norm": 0.43359375,
      "learning_rate": 4.660024111137167e-06,
      "loss": 0.4987,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5231
    },
    {
      "epoch": 0.7114495512646178,
      "grad_norm": 0.392578125,
      "learning_rate": 4.656010548704574e-06,
      "loss": 0.653,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5232
    },
    {
      "epoch": 0.7115855316834376,
      "grad_norm": 0.625,
      "learning_rate": 4.651998190872537e-06,
      "loss": 0.5653,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5233
    },
    {
      "epoch": 0.7117215121022573,
      "grad_norm": 0.310546875,
      "learning_rate": 4.647987038545496e-06,
      "loss": 0.6663,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5234
    },
    {
      "epoch": 0.711857492521077,
      "grad_norm": 0.3984375,
      "learning_rate": 4.643977092627608e-06,
      "loss": 0.7498,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5235
    },
    {
      "epoch": 0.7119934729398967,
      "grad_norm": 0.353515625,
      "learning_rate": 4.639968354022768e-06,
      "loss": 0.6367,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5236
    },
    {
      "epoch": 0.7121294533587164,
      "grad_norm": 0.388671875,
      "learning_rate": 4.635960823634591e-06,
      "loss": 0.7674,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5237
    },
    {
      "epoch": 0.712265433777536,
      "grad_norm": 0.640625,
      "learning_rate": 4.631954502366428e-06,
      "loss": 0.6532,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5238
    },
    {
      "epoch": 0.7124014141963557,
      "grad_norm": 0.380859375,
      "learning_rate": 4.627949391121348e-06,
      "loss": 0.6772,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5239
    },
    {
      "epoch": 0.7125373946151754,
      "grad_norm": 0.34765625,
      "learning_rate": 4.6239454908021575e-06,
      "loss": 0.6605,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5240
    },
    {
      "epoch": 0.7126733750339951,
      "grad_norm": 0.35546875,
      "learning_rate": 4.619942802311378e-06,
      "loss": 0.6347,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5241
    },
    {
      "epoch": 0.7128093554528148,
      "grad_norm": 0.431640625,
      "learning_rate": 4.6159413265512706e-06,
      "loss": 0.6299,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5242
    },
    {
      "epoch": 0.7129453358716344,
      "grad_norm": 0.69140625,
      "learning_rate": 4.611941064423812e-06,
      "loss": 0.7446,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5243
    },
    {
      "epoch": 0.7130813162904541,
      "grad_norm": 0.337890625,
      "learning_rate": 4.607942016830714e-06,
      "loss": 0.619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5244
    },
    {
      "epoch": 0.7132172967092739,
      "grad_norm": 0.37890625,
      "learning_rate": 4.603944184673407e-06,
      "loss": 0.7594,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5245
    },
    {
      "epoch": 0.7133532771280936,
      "grad_norm": 0.74609375,
      "learning_rate": 4.5999475688530574e-06,
      "loss": 0.7212,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5246
    },
    {
      "epoch": 0.7134892575469133,
      "grad_norm": 0.251953125,
      "learning_rate": 4.595952170270542e-06,
      "loss": 0.4507,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5247
    },
    {
      "epoch": 0.713625237965733,
      "grad_norm": 0.376953125,
      "learning_rate": 4.5919579898264824e-06,
      "loss": 0.7181,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5248
    },
    {
      "epoch": 0.7137612183845526,
      "grad_norm": 0.474609375,
      "learning_rate": 4.587965028421207e-06,
      "loss": 0.423,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5249
    },
    {
      "epoch": 0.7138971988033723,
      "grad_norm": 0.423828125,
      "learning_rate": 4.583973286954786e-06,
      "loss": 0.6488,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5250
    },
    {
      "epoch": 0.714033179222192,
      "grad_norm": 0.53515625,
      "learning_rate": 4.579982766326999e-06,
      "loss": 0.7887,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5251
    },
    {
      "epoch": 0.7141691596410117,
      "grad_norm": 0.40234375,
      "learning_rate": 4.575993467437366e-06,
      "loss": 0.6325,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5252
    },
    {
      "epoch": 0.7143051400598314,
      "grad_norm": 0.392578125,
      "learning_rate": 4.572005391185117e-06,
      "loss": 0.7729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5253
    },
    {
      "epoch": 0.714441120478651,
      "grad_norm": 0.64453125,
      "learning_rate": 4.568018538469219e-06,
      "loss": 0.6729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5254
    },
    {
      "epoch": 0.7145771008974707,
      "grad_norm": 0.4765625,
      "learning_rate": 4.56403291018835e-06,
      "loss": 0.6719,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5255
    },
    {
      "epoch": 0.7147130813162904,
      "grad_norm": 0.609375,
      "learning_rate": 4.560048507240927e-06,
      "loss": 0.699,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5256
    },
    {
      "epoch": 0.7148490617351101,
      "grad_norm": 0.4140625,
      "learning_rate": 4.556065330525077e-06,
      "loss": 0.6289,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5257
    },
    {
      "epoch": 0.7149850421539299,
      "grad_norm": 0.625,
      "learning_rate": 4.552083380938662e-06,
      "loss": 0.9701,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5258
    },
    {
      "epoch": 0.7151210225727496,
      "grad_norm": 0.310546875,
      "learning_rate": 4.5481026593792574e-06,
      "loss": 0.619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5259
    },
    {
      "epoch": 0.7152570029915692,
      "grad_norm": 0.435546875,
      "learning_rate": 4.5441231667441724e-06,
      "loss": 0.7148,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5260
    },
    {
      "epoch": 0.7153929834103889,
      "grad_norm": 0.345703125,
      "learning_rate": 4.540144903930425e-06,
      "loss": 0.6442,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5261
    },
    {
      "epoch": 0.7155289638292086,
      "grad_norm": 0.33984375,
      "learning_rate": 4.536167871834772e-06,
      "loss": 0.6252,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5262
    },
    {
      "epoch": 0.7156649442480283,
      "grad_norm": 0.4140625,
      "learning_rate": 4.53219207135368e-06,
      "loss": 0.6502,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5263
    },
    {
      "epoch": 0.715800924666848,
      "grad_norm": 0.29296875,
      "learning_rate": 4.5282175033833465e-06,
      "loss": 0.4564,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5264
    },
    {
      "epoch": 0.7159369050856677,
      "grad_norm": 0.330078125,
      "learning_rate": 4.524244168819684e-06,
      "loss": 0.616,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5265
    },
    {
      "epoch": 0.7160728855044873,
      "grad_norm": 0.30078125,
      "learning_rate": 4.520272068558337e-06,
      "loss": 0.5177,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5266
    },
    {
      "epoch": 0.716208865923307,
      "grad_norm": 0.392578125,
      "learning_rate": 4.516301203494656e-06,
      "loss": 0.6444,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5267
    },
    {
      "epoch": 0.7163448463421267,
      "grad_norm": 0.2890625,
      "learning_rate": 4.512331574523733e-06,
      "loss": 0.5422,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5268
    },
    {
      "epoch": 0.7164808267609464,
      "grad_norm": 0.73828125,
      "learning_rate": 4.508363182540361e-06,
      "loss": 0.7051,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5269
    },
    {
      "epoch": 0.7166168071797662,
      "grad_norm": 0.39453125,
      "learning_rate": 4.504396028439073e-06,
      "loss": 0.7228,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5270
    },
    {
      "epoch": 0.7167527875985858,
      "grad_norm": 0.341796875,
      "learning_rate": 4.50043011311411e-06,
      "loss": 0.6126,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5271
    },
    {
      "epoch": 0.7168887680174055,
      "grad_norm": 0.490234375,
      "learning_rate": 4.496465437459434e-06,
      "loss": 0.6735,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5272
    },
    {
      "epoch": 0.7170247484362252,
      "grad_norm": 0.35546875,
      "learning_rate": 4.492502002368738e-06,
      "loss": 0.6628,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5273
    },
    {
      "epoch": 0.7171607288550449,
      "grad_norm": 0.451171875,
      "learning_rate": 4.4885398087354235e-06,
      "loss": 0.8301,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5274
    },
    {
      "epoch": 0.7172967092738646,
      "grad_norm": 0.234375,
      "learning_rate": 4.484578857452623e-06,
      "loss": 0.4027,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5275
    },
    {
      "epoch": 0.7174326896926843,
      "grad_norm": 0.28125,
      "learning_rate": 4.480619149413176e-06,
      "loss": 0.5674,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5276
    },
    {
      "epoch": 0.7175686701115039,
      "grad_norm": 0.26171875,
      "learning_rate": 4.476660685509656e-06,
      "loss": 0.4883,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5277
    },
    {
      "epoch": 0.7177046505303236,
      "grad_norm": 0.59375,
      "learning_rate": 4.47270346663434e-06,
      "loss": 0.7471,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5278
    },
    {
      "epoch": 0.7178406309491433,
      "grad_norm": 0.341796875,
      "learning_rate": 4.468747493679246e-06,
      "loss": 0.7183,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5279
    },
    {
      "epoch": 0.717976611367963,
      "grad_norm": 0.369140625,
      "learning_rate": 4.46479276753609e-06,
      "loss": 0.6341,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5280
    },
    {
      "epoch": 0.7181125917867827,
      "grad_norm": 0.361328125,
      "learning_rate": 4.4608392890963214e-06,
      "loss": 0.7144,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5281
    },
    {
      "epoch": 0.7182485722056023,
      "grad_norm": 0.416015625,
      "learning_rate": 4.456887059251096e-06,
      "loss": 0.7407,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5282
    },
    {
      "epoch": 0.7183845526244221,
      "grad_norm": 0.38671875,
      "learning_rate": 4.452936078891301e-06,
      "loss": 0.7352,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5283
    },
    {
      "epoch": 0.7185205330432418,
      "grad_norm": 0.345703125,
      "learning_rate": 4.448986348907529e-06,
      "loss": 0.6074,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5284
    },
    {
      "epoch": 0.7186565134620615,
      "grad_norm": 0.37109375,
      "learning_rate": 4.445037870190106e-06,
      "loss": 0.7324,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5285
    },
    {
      "epoch": 0.7187924938808812,
      "grad_norm": 0.29296875,
      "learning_rate": 4.441090643629057e-06,
      "loss": 0.59,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5286
    },
    {
      "epoch": 0.7189284742997009,
      "grad_norm": 0.4140625,
      "learning_rate": 4.437144670114145e-06,
      "loss": 0.7249,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5287
    },
    {
      "epoch": 0.7190644547185205,
      "grad_norm": 0.328125,
      "learning_rate": 4.433199950534832e-06,
      "loss": 0.7008,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5288
    },
    {
      "epoch": 0.7192004351373402,
      "grad_norm": 0.271484375,
      "learning_rate": 4.429256485780314e-06,
      "loss": 0.5081,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5289
    },
    {
      "epoch": 0.7193364155561599,
      "grad_norm": 0.53125,
      "learning_rate": 4.42531427673949e-06,
      "loss": 0.7943,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5290
    },
    {
      "epoch": 0.7194723959749796,
      "grad_norm": 0.33203125,
      "learning_rate": 4.421373324300988e-06,
      "loss": 0.5036,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5291
    },
    {
      "epoch": 0.7196083763937993,
      "grad_norm": 0.51171875,
      "learning_rate": 4.417433629353139e-06,
      "loss": 0.6842,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5292
    },
    {
      "epoch": 0.719744356812619,
      "grad_norm": 0.392578125,
      "learning_rate": 4.4134951927840074e-06,
      "loss": 0.6481,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5293
    },
    {
      "epoch": 0.7198803372314386,
      "grad_norm": 0.291015625,
      "learning_rate": 4.409558015481355e-06,
      "loss": 0.5246,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5294
    },
    {
      "epoch": 0.7200163176502583,
      "grad_norm": 0.462890625,
      "learning_rate": 4.40562209833268e-06,
      "loss": 0.5299,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5295
    },
    {
      "epoch": 0.7201522980690781,
      "grad_norm": 0.5078125,
      "learning_rate": 4.401687442225179e-06,
      "loss": 0.7679,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5296
    },
    {
      "epoch": 0.7202882784878978,
      "grad_norm": 0.70703125,
      "learning_rate": 4.397754048045776e-06,
      "loss": 0.6572,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5297
    },
    {
      "epoch": 0.7204242589067175,
      "grad_norm": 0.47265625,
      "learning_rate": 4.3938219166811005e-06,
      "loss": 0.7358,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5298
    },
    {
      "epoch": 0.7205602393255371,
      "grad_norm": 0.546875,
      "learning_rate": 4.389891049017511e-06,
      "loss": 0.8188,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5299
    },
    {
      "epoch": 0.7206962197443568,
      "grad_norm": 0.466796875,
      "learning_rate": 4.385961445941064e-06,
      "loss": 0.8861,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5300
    },
    {
      "epoch": 0.7208322001631765,
      "grad_norm": 0.3125,
      "learning_rate": 4.3820331083375485e-06,
      "loss": 0.6405,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5301
    },
    {
      "epoch": 0.7209681805819962,
      "grad_norm": 0.267578125,
      "learning_rate": 4.378106037092452e-06,
      "loss": 0.377,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5302
    },
    {
      "epoch": 0.7211041610008159,
      "grad_norm": 0.61328125,
      "learning_rate": 4.3741802330909935e-06,
      "loss": 0.7637,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5303
    },
    {
      "epoch": 0.7212401414196355,
      "grad_norm": 0.431640625,
      "learning_rate": 4.370255697218087e-06,
      "loss": 0.6582,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5304
    },
    {
      "epoch": 0.7213761218384552,
      "grad_norm": 0.4765625,
      "learning_rate": 4.366332430358382e-06,
      "loss": 0.5967,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5305
    },
    {
      "epoch": 0.7215121022572749,
      "grad_norm": 0.57421875,
      "learning_rate": 4.362410433396219e-06,
      "loss": 0.6159,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5306
    },
    {
      "epoch": 0.7216480826760946,
      "grad_norm": 0.68359375,
      "learning_rate": 4.358489707215675e-06,
      "loss": 0.7979,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5307
    },
    {
      "epoch": 0.7217840630949144,
      "grad_norm": 0.412109375,
      "learning_rate": 4.35457025270052e-06,
      "loss": 0.6572,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5308
    },
    {
      "epoch": 0.7219200435137341,
      "grad_norm": 0.369140625,
      "learning_rate": 4.350652070734256e-06,
      "loss": 0.7539,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5309
    },
    {
      "epoch": 0.7220560239325537,
      "grad_norm": 0.423828125,
      "learning_rate": 4.34673516220008e-06,
      "loss": 0.5902,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5310
    },
    {
      "epoch": 0.7221920043513734,
      "grad_norm": 0.35546875,
      "learning_rate": 4.34281952798092e-06,
      "loss": 0.6377,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5311
    },
    {
      "epoch": 0.7223279847701931,
      "grad_norm": 0.44140625,
      "learning_rate": 4.3389051689594e-06,
      "loss": 0.4766,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5312
    },
    {
      "epoch": 0.7224639651890128,
      "grad_norm": 0.546875,
      "learning_rate": 4.334992086017871e-06,
      "loss": 0.6086,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5313
    },
    {
      "epoch": 0.7225999456078325,
      "grad_norm": 0.33984375,
      "learning_rate": 4.331080280038384e-06,
      "loss": 0.7114,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5314
    },
    {
      "epoch": 0.7227359260266522,
      "grad_norm": 0.375,
      "learning_rate": 4.3271697519027126e-06,
      "loss": 0.7528,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5315
    },
    {
      "epoch": 0.7228719064454718,
      "grad_norm": 0.57421875,
      "learning_rate": 4.323260502492333e-06,
      "loss": 0.5316,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5316
    },
    {
      "epoch": 0.7230078868642915,
      "grad_norm": 0.412109375,
      "learning_rate": 4.319352532688444e-06,
      "loss": 0.6211,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5317
    },
    {
      "epoch": 0.7231438672831112,
      "grad_norm": 0.197265625,
      "learning_rate": 4.315445843371941e-06,
      "loss": 0.3877,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5318
    },
    {
      "epoch": 0.7232798477019309,
      "grad_norm": 0.412109375,
      "learning_rate": 4.31154043542345e-06,
      "loss": 0.6192,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5319
    },
    {
      "epoch": 0.7234158281207506,
      "grad_norm": 0.4140625,
      "learning_rate": 4.307636309723289e-06,
      "loss": 0.7134,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5320
    },
    {
      "epoch": 0.7235518085395704,
      "grad_norm": 0.373046875,
      "learning_rate": 4.303733467151503e-06,
      "loss": 0.6086,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5321
    },
    {
      "epoch": 0.72368778895839,
      "grad_norm": 0.46484375,
      "learning_rate": 4.2998319085878316e-06,
      "loss": 0.6836,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5322
    },
    {
      "epoch": 0.7238237693772097,
      "grad_norm": 0.2734375,
      "learning_rate": 4.295931634911743e-06,
      "loss": 0.5868,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5323
    },
    {
      "epoch": 0.7239597497960294,
      "grad_norm": 0.57421875,
      "learning_rate": 4.292032647002402e-06,
      "loss": 0.7301,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5324
    },
    {
      "epoch": 0.7240957302148491,
      "grad_norm": 0.58984375,
      "learning_rate": 4.288134945738684e-06,
      "loss": 0.7196,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5325
    },
    {
      "epoch": 0.7242317106336688,
      "grad_norm": 0.3671875,
      "learning_rate": 4.2842385319991886e-06,
      "loss": 0.418,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5326
    },
    {
      "epoch": 0.7243676910524884,
      "grad_norm": 0.2275390625,
      "learning_rate": 4.280343406662205e-06,
      "loss": 0.4302,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5327
    },
    {
      "epoch": 0.7245036714713081,
      "grad_norm": 0.2890625,
      "learning_rate": 4.27644957060575e-06,
      "loss": 0.5356,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5328
    },
    {
      "epoch": 0.7246396518901278,
      "grad_norm": 0.3515625,
      "learning_rate": 4.272557024707535e-06,
      "loss": 0.7111,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5329
    },
    {
      "epoch": 0.7247756323089475,
      "grad_norm": 0.296875,
      "learning_rate": 4.2686657698449926e-06,
      "loss": 0.5716,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5330
    },
    {
      "epoch": 0.7249116127277672,
      "grad_norm": 0.390625,
      "learning_rate": 4.264775806895256e-06,
      "loss": 0.7386,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5331
    },
    {
      "epoch": 0.7250475931465868,
      "grad_norm": 0.458984375,
      "learning_rate": 4.260887136735175e-06,
      "loss": 0.6528,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5332
    },
    {
      "epoch": 0.7251835735654066,
      "grad_norm": 0.283203125,
      "learning_rate": 4.256999760241296e-06,
      "loss": 0.4813,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5333
    },
    {
      "epoch": 0.7253195539842263,
      "grad_norm": 0.482421875,
      "learning_rate": 4.253113678289889e-06,
      "loss": 0.6444,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5334
    },
    {
      "epoch": 0.725455534403046,
      "grad_norm": 0.470703125,
      "learning_rate": 4.249228891756918e-06,
      "loss": 0.8034,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5335
    },
    {
      "epoch": 0.7255915148218657,
      "grad_norm": 0.369140625,
      "learning_rate": 4.245345401518067e-06,
      "loss": 0.6937,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5336
    },
    {
      "epoch": 0.7257274952406854,
      "grad_norm": 0.494140625,
      "learning_rate": 4.241463208448716e-06,
      "loss": 0.6136,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5337
    },
    {
      "epoch": 0.725863475659505,
      "grad_norm": 0.376953125,
      "learning_rate": 4.2375823134239624e-06,
      "loss": 0.6089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5338
    },
    {
      "epoch": 0.7259994560783247,
      "grad_norm": 0.279296875,
      "learning_rate": 4.2337027173186074e-06,
      "loss": 0.5793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5339
    },
    {
      "epoch": 0.7261354364971444,
      "grad_norm": 0.6171875,
      "learning_rate": 4.229824421007162e-06,
      "loss": 0.7464,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5340
    },
    {
      "epoch": 0.7262714169159641,
      "grad_norm": 0.796875,
      "learning_rate": 4.225947425363835e-06,
      "loss": 0.666,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5341
    },
    {
      "epoch": 0.7264073973347838,
      "grad_norm": 0.33984375,
      "learning_rate": 4.222071731262557e-06,
      "loss": 0.6889,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5342
    },
    {
      "epoch": 0.7265433777536034,
      "grad_norm": 0.5390625,
      "learning_rate": 4.218197339576948e-06,
      "loss": 0.8545,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5343
    },
    {
      "epoch": 0.7266793581724231,
      "grad_norm": 0.359375,
      "learning_rate": 4.214324251180351e-06,
      "loss": 0.784,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5344
    },
    {
      "epoch": 0.7268153385912428,
      "grad_norm": 0.2431640625,
      "learning_rate": 4.2104524669458e-06,
      "loss": 0.4022,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5345
    },
    {
      "epoch": 0.7269513190100626,
      "grad_norm": 0.359375,
      "learning_rate": 4.20658198774605e-06,
      "loss": 0.7036,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5346
    },
    {
      "epoch": 0.7270872994288823,
      "grad_norm": 0.58984375,
      "learning_rate": 4.202712814453548e-06,
      "loss": 0.6644,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5347
    },
    {
      "epoch": 0.727223279847702,
      "grad_norm": 0.265625,
      "learning_rate": 4.198844947940459e-06,
      "loss": 0.5073,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5348
    },
    {
      "epoch": 0.7273592602665216,
      "grad_norm": 0.625,
      "learning_rate": 4.19497838907864e-06,
      "loss": 0.8436,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5349
    },
    {
      "epoch": 0.7274952406853413,
      "grad_norm": 0.31640625,
      "learning_rate": 4.1911131387396686e-06,
      "loss": 0.4425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5350
    },
    {
      "epoch": 0.727631221104161,
      "grad_norm": 0.33984375,
      "learning_rate": 4.1872491977948125e-06,
      "loss": 0.6043,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5351
    },
    {
      "epoch": 0.7277672015229807,
      "grad_norm": 0.35546875,
      "learning_rate": 4.183386567115057e-06,
      "loss": 0.5301,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5352
    },
    {
      "epoch": 0.7279031819418004,
      "grad_norm": 0.45703125,
      "learning_rate": 4.179525247571082e-06,
      "loss": 0.7186,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5353
    },
    {
      "epoch": 0.72803916236062,
      "grad_norm": 0.404296875,
      "learning_rate": 4.175665240033279e-06,
      "loss": 0.6721,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5354
    },
    {
      "epoch": 0.7281751427794397,
      "grad_norm": 0.29296875,
      "learning_rate": 4.171806545371737e-06,
      "loss": 0.4909,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5355
    },
    {
      "epoch": 0.7283111231982594,
      "grad_norm": 0.486328125,
      "learning_rate": 4.16794916445626e-06,
      "loss": 0.4673,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5356
    },
    {
      "epoch": 0.7284471036170791,
      "grad_norm": 0.333984375,
      "learning_rate": 4.16409309815634e-06,
      "loss": 0.6413,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5357
    },
    {
      "epoch": 0.7285830840358988,
      "grad_norm": 0.296875,
      "learning_rate": 4.160238347341189e-06,
      "loss": 0.6235,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5358
    },
    {
      "epoch": 0.7287190644547186,
      "grad_norm": 0.341796875,
      "learning_rate": 4.15638491287971e-06,
      "loss": 0.6385,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5359
    },
    {
      "epoch": 0.7288550448735382,
      "grad_norm": 0.412109375,
      "learning_rate": 4.152532795640518e-06,
      "loss": 0.6286,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5360
    },
    {
      "epoch": 0.7289910252923579,
      "grad_norm": 0.33203125,
      "learning_rate": 4.148681996491923e-06,
      "loss": 0.5851,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5361
    },
    {
      "epoch": 0.7291270057111776,
      "grad_norm": 0.38671875,
      "learning_rate": 4.144832516301947e-06,
      "loss": 0.7114,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5362
    },
    {
      "epoch": 0.7292629861299973,
      "grad_norm": 0.349609375,
      "learning_rate": 4.140984355938306e-06,
      "loss": 0.6696,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5363
    },
    {
      "epoch": 0.729398966548817,
      "grad_norm": 0.400390625,
      "learning_rate": 4.1371375162684255e-06,
      "loss": 0.7324,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5364
    },
    {
      "epoch": 0.7295349469676367,
      "grad_norm": 0.369140625,
      "learning_rate": 4.133291998159427e-06,
      "loss": 0.6517,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5365
    },
    {
      "epoch": 0.7296709273864563,
      "grad_norm": 0.38671875,
      "learning_rate": 4.129447802478142e-06,
      "loss": 0.6829,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5366
    },
    {
      "epoch": 0.729806907805276,
      "grad_norm": 0.2392578125,
      "learning_rate": 4.1256049300910914e-06,
      "loss": 0.361,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5367
    },
    {
      "epoch": 0.7299428882240957,
      "grad_norm": 0.33984375,
      "learning_rate": 4.121763381864514e-06,
      "loss": 0.5924,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5368
    },
    {
      "epoch": 0.7300788686429154,
      "grad_norm": 0.353515625,
      "learning_rate": 4.117923158664335e-06,
      "loss": 0.7096,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5369
    },
    {
      "epoch": 0.7302148490617351,
      "grad_norm": 0.263671875,
      "learning_rate": 4.114084261356195e-06,
      "loss": 0.5617,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5370
    },
    {
      "epoch": 0.7303508294805549,
      "grad_norm": 0.380859375,
      "learning_rate": 4.1102466908054206e-06,
      "loss": 0.7288,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5371
    },
    {
      "epoch": 0.7304868098993745,
      "grad_norm": 0.328125,
      "learning_rate": 4.106410447877054e-06,
      "loss": 0.7127,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5372
    },
    {
      "epoch": 0.7306227903181942,
      "grad_norm": 0.640625,
      "learning_rate": 4.102575533435824e-06,
      "loss": 0.6089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5373
    },
    {
      "epoch": 0.7307587707370139,
      "grad_norm": 0.60546875,
      "learning_rate": 4.098741948346175e-06,
      "loss": 0.4271,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5374
    },
    {
      "epoch": 0.7308947511558336,
      "grad_norm": 0.423828125,
      "learning_rate": 4.094909693472237e-06,
      "loss": 0.6429,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5375
    },
    {
      "epoch": 0.7310307315746533,
      "grad_norm": 0.71484375,
      "learning_rate": 4.091078769677852e-06,
      "loss": 0.7116,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5376
    },
    {
      "epoch": 0.7311667119934729,
      "grad_norm": 0.5546875,
      "learning_rate": 4.087249177826553e-06,
      "loss": 0.742,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5377
    },
    {
      "epoch": 0.7313026924122926,
      "grad_norm": 0.36328125,
      "learning_rate": 4.083420918781583e-06,
      "loss": 0.6488,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5378
    },
    {
      "epoch": 0.7314386728311123,
      "grad_norm": 0.349609375,
      "learning_rate": 4.079593993405874e-06,
      "loss": 0.6315,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5379
    },
    {
      "epoch": 0.731574653249932,
      "grad_norm": 0.375,
      "learning_rate": 4.075768402562059e-06,
      "loss": 0.6126,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5380
    },
    {
      "epoch": 0.7317106336687517,
      "grad_norm": 0.255859375,
      "learning_rate": 4.0719441471124784e-06,
      "loss": 0.4806,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5381
    },
    {
      "epoch": 0.7318466140875713,
      "grad_norm": 0.62109375,
      "learning_rate": 4.068121227919162e-06,
      "loss": 0.8241,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5382
    },
    {
      "epoch": 0.731982594506391,
      "grad_norm": 0.4296875,
      "learning_rate": 4.064299645843848e-06,
      "loss": 0.7689,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5383
    },
    {
      "epoch": 0.7321185749252108,
      "grad_norm": 1.2734375,
      "learning_rate": 4.06047940174796e-06,
      "loss": 0.8557,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5384
    },
    {
      "epoch": 0.7322545553440305,
      "grad_norm": 0.345703125,
      "learning_rate": 4.056660496492635e-06,
      "loss": 0.7357,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5385
    },
    {
      "epoch": 0.7323905357628502,
      "grad_norm": 0.404296875,
      "learning_rate": 4.0528429309386955e-06,
      "loss": 0.4222,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5386
    },
    {
      "epoch": 0.7325265161816699,
      "grad_norm": 0.490234375,
      "learning_rate": 4.0490267059466714e-06,
      "loss": 0.8638,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5387
    },
    {
      "epoch": 0.7326624966004895,
      "grad_norm": 0.67578125,
      "learning_rate": 4.045211822376784e-06,
      "loss": 0.7264,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5388
    },
    {
      "epoch": 0.7327984770193092,
      "grad_norm": 0.314453125,
      "learning_rate": 4.041398281088956e-06,
      "loss": 0.6037,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5389
    },
    {
      "epoch": 0.7329344574381289,
      "grad_norm": 0.5546875,
      "learning_rate": 4.037586082942805e-06,
      "loss": 0.7954,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5390
    },
    {
      "epoch": 0.7330704378569486,
      "grad_norm": 0.44140625,
      "learning_rate": 4.03377522879765e-06,
      "loss": 0.6485,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5391
    },
    {
      "epoch": 0.7332064182757683,
      "grad_norm": 0.6796875,
      "learning_rate": 4.029965719512498e-06,
      "loss": 0.8582,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5392
    },
    {
      "epoch": 0.733342398694588,
      "grad_norm": 0.2578125,
      "learning_rate": 4.026157555946067e-06,
      "loss": 0.5422,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5393
    },
    {
      "epoch": 0.7334783791134076,
      "grad_norm": 0.3515625,
      "learning_rate": 4.022350738956756e-06,
      "loss": 0.6264,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5394
    },
    {
      "epoch": 0.7336143595322273,
      "grad_norm": 0.333984375,
      "learning_rate": 4.0185452694026764e-06,
      "loss": 0.4697,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5395
    },
    {
      "epoch": 0.7337503399510471,
      "grad_norm": 0.416015625,
      "learning_rate": 4.014741148141618e-06,
      "loss": 0.7368,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5396
    },
    {
      "epoch": 0.7338863203698668,
      "grad_norm": 0.3359375,
      "learning_rate": 4.010938376031085e-06,
      "loss": 0.6084,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5397
    },
    {
      "epoch": 0.7340223007886865,
      "grad_norm": 0.4296875,
      "learning_rate": 4.0071369539282565e-06,
      "loss": 0.6891,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5398
    },
    {
      "epoch": 0.7341582812075061,
      "grad_norm": 0.76171875,
      "learning_rate": 4.003336882690036e-06,
      "loss": 0.5326,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5399
    },
    {
      "epoch": 0.7342942616263258,
      "grad_norm": 0.4375,
      "learning_rate": 3.999538163172995e-06,
      "loss": 0.6893,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5400
    },
    {
      "epoch": 0.7344302420451455,
      "grad_norm": 0.328125,
      "learning_rate": 3.995740796233416e-06,
      "loss": 0.6805,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5401
    },
    {
      "epoch": 0.7345662224639652,
      "grad_norm": 0.54296875,
      "learning_rate": 3.991944782727268e-06,
      "loss": 0.5806,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5402
    },
    {
      "epoch": 0.7347022028827849,
      "grad_norm": 0.296875,
      "learning_rate": 3.988150123510224e-06,
      "loss": 0.5369,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5403
    },
    {
      "epoch": 0.7348381833016046,
      "grad_norm": 0.408203125,
      "learning_rate": 3.98435681943764e-06,
      "loss": 0.7695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5404
    },
    {
      "epoch": 0.7349741637204242,
      "grad_norm": 0.5703125,
      "learning_rate": 3.98056487136458e-06,
      "loss": 0.6998,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5405
    },
    {
      "epoch": 0.7351101441392439,
      "grad_norm": 0.29296875,
      "learning_rate": 3.976774280145788e-06,
      "loss": 0.5653,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5406
    },
    {
      "epoch": 0.7352461245580636,
      "grad_norm": 0.5546875,
      "learning_rate": 3.972985046635718e-06,
      "loss": 0.6774,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5407
    },
    {
      "epoch": 0.7353821049768833,
      "grad_norm": 0.76171875,
      "learning_rate": 3.9691971716885015e-06,
      "loss": 0.855,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5408
    },
    {
      "epoch": 0.7355180853957031,
      "grad_norm": 0.57421875,
      "learning_rate": 3.96541065615798e-06,
      "loss": 0.5515,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5409
    },
    {
      "epoch": 0.7356540658145228,
      "grad_norm": 0.376953125,
      "learning_rate": 3.961625500897671e-06,
      "loss": 0.5907,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5410
    },
    {
      "epoch": 0.7357900462333424,
      "grad_norm": 1.84375,
      "learning_rate": 3.957841706760804e-06,
      "loss": 0.8464,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5411
    },
    {
      "epoch": 0.7359260266521621,
      "grad_norm": 0.443359375,
      "learning_rate": 3.954059274600285e-06,
      "loss": 0.7702,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5412
    },
    {
      "epoch": 0.7360620070709818,
      "grad_norm": 0.64453125,
      "learning_rate": 3.950278205268726e-06,
      "loss": 0.5182,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5413
    },
    {
      "epoch": 0.7361979874898015,
      "grad_norm": 0.310546875,
      "learning_rate": 3.9464984996184204e-06,
      "loss": 0.6452,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5414
    },
    {
      "epoch": 0.7363339679086212,
      "grad_norm": 0.474609375,
      "learning_rate": 3.942720158501369e-06,
      "loss": 0.7687,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5415
    },
    {
      "epoch": 0.7364699483274408,
      "grad_norm": 1.1953125,
      "learning_rate": 3.9389431827692455e-06,
      "loss": 0.7656,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5416
    },
    {
      "epoch": 0.7366059287462605,
      "grad_norm": 0.52734375,
      "learning_rate": 3.935167573273436e-06,
      "loss": 0.5579,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5417
    },
    {
      "epoch": 0.7367419091650802,
      "grad_norm": 0.68359375,
      "learning_rate": 3.931393330865e-06,
      "loss": 0.9152,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5418
    },
    {
      "epoch": 0.7368778895838999,
      "grad_norm": 0.34375,
      "learning_rate": 3.927620456394707e-06,
      "loss": 0.5529,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5419
    },
    {
      "epoch": 0.7370138700027196,
      "grad_norm": 0.400390625,
      "learning_rate": 3.9238489507130005e-06,
      "loss": 0.8148,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5420
    },
    {
      "epoch": 0.7371498504215392,
      "grad_norm": 0.310546875,
      "learning_rate": 3.920078814670032e-06,
      "loss": 0.5422,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5421
    },
    {
      "epoch": 0.737285830840359,
      "grad_norm": 0.333984375,
      "learning_rate": 3.916310049115628e-06,
      "loss": 0.5899,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5422
    },
    {
      "epoch": 0.7374218112591787,
      "grad_norm": 0.337890625,
      "learning_rate": 3.912542654899323e-06,
      "loss": 0.5975,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5423
    },
    {
      "epoch": 0.7375577916779984,
      "grad_norm": 0.3359375,
      "learning_rate": 3.908776632870324e-06,
      "loss": 0.5941,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5424
    },
    {
      "epoch": 0.7376937720968181,
      "grad_norm": 0.72265625,
      "learning_rate": 3.9050119838775505e-06,
      "loss": 0.8716,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5425
    },
    {
      "epoch": 0.7378297525156378,
      "grad_norm": 0.333984375,
      "learning_rate": 3.901248708769587e-06,
      "loss": 0.6296,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5426
    },
    {
      "epoch": 0.7379657329344574,
      "grad_norm": 0.306640625,
      "learning_rate": 3.897486808394733e-06,
      "loss": 0.5251,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5427
    },
    {
      "epoch": 0.7381017133532771,
      "grad_norm": 0.38671875,
      "learning_rate": 3.893726283600959e-06,
      "loss": 0.6629,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5428
    },
    {
      "epoch": 0.7382376937720968,
      "grad_norm": 0.56640625,
      "learning_rate": 3.88996713523594e-06,
      "loss": 0.6063,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5429
    },
    {
      "epoch": 0.7383736741909165,
      "grad_norm": 0.55859375,
      "learning_rate": 3.886209364147027e-06,
      "loss": 0.8604,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5430
    },
    {
      "epoch": 0.7385096546097362,
      "grad_norm": 0.2177734375,
      "learning_rate": 3.882452971181274e-06,
      "loss": 0.3597,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5431
    },
    {
      "epoch": 0.7386456350285558,
      "grad_norm": 0.51953125,
      "learning_rate": 3.8786979571854154e-06,
      "loss": 0.8896,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5432
    },
    {
      "epoch": 0.7387816154473755,
      "grad_norm": 0.265625,
      "learning_rate": 3.874944323005873e-06,
      "loss": 0.4513,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5433
    },
    {
      "epoch": 0.7389175958661953,
      "grad_norm": 0.47265625,
      "learning_rate": 3.87119206948877e-06,
      "loss": 0.7697,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5434
    },
    {
      "epoch": 0.739053576285015,
      "grad_norm": 0.330078125,
      "learning_rate": 3.867441197479901e-06,
      "loss": 0.577,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5435
    },
    {
      "epoch": 0.7391895567038347,
      "grad_norm": 0.35546875,
      "learning_rate": 3.8636917078247675e-06,
      "loss": 0.5163,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5436
    },
    {
      "epoch": 0.7393255371226544,
      "grad_norm": 0.39453125,
      "learning_rate": 3.859943601368542e-06,
      "loss": 0.7633,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5437
    },
    {
      "epoch": 0.739461517541474,
      "grad_norm": 0.326171875,
      "learning_rate": 3.856196878956101e-06,
      "loss": 0.716,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5438
    },
    {
      "epoch": 0.7395974979602937,
      "grad_norm": 0.33203125,
      "learning_rate": 3.8524515414319955e-06,
      "loss": 0.2051,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5439
    },
    {
      "epoch": 0.7397334783791134,
      "grad_norm": 0.33203125,
      "learning_rate": 3.848707589640476e-06,
      "loss": 0.5706,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5440
    },
    {
      "epoch": 0.7398694587979331,
      "grad_norm": 0.365234375,
      "learning_rate": 3.844965024425468e-06,
      "loss": 0.5975,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5441
    },
    {
      "epoch": 0.7400054392167528,
      "grad_norm": 0.396484375,
      "learning_rate": 3.841223846630599e-06,
      "loss": 0.7401,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5442
    },
    {
      "epoch": 0.7401414196355725,
      "grad_norm": 0.62890625,
      "learning_rate": 3.83748405709917e-06,
      "loss": 0.6221,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5443
    },
    {
      "epoch": 0.7402774000543921,
      "grad_norm": 0.283203125,
      "learning_rate": 3.833745656674181e-06,
      "loss": 0.5179,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5444
    },
    {
      "epoch": 0.7404133804732118,
      "grad_norm": 0.349609375,
      "learning_rate": 3.8300086461983065e-06,
      "loss": 0.5918,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5445
    },
    {
      "epoch": 0.7405493608920315,
      "grad_norm": 0.4921875,
      "learning_rate": 3.8262730265139215e-06,
      "loss": 0.8346,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5446
    },
    {
      "epoch": 0.7406853413108513,
      "grad_norm": 0.310546875,
      "learning_rate": 3.8225387984630735e-06,
      "loss": 0.6553,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5447
    },
    {
      "epoch": 0.740821321729671,
      "grad_norm": 0.79296875,
      "learning_rate": 3.818805962887509e-06,
      "loss": 0.8037,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5448
    },
    {
      "epoch": 0.7409573021484906,
      "grad_norm": 0.39453125,
      "learning_rate": 3.81507452062865e-06,
      "loss": 0.6489,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5449
    },
    {
      "epoch": 0.7410932825673103,
      "grad_norm": 0.380859375,
      "learning_rate": 3.811344472527615e-06,
      "loss": 0.7347,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5450
    },
    {
      "epoch": 0.74122926298613,
      "grad_norm": 0.3203125,
      "learning_rate": 3.8076158194251945e-06,
      "loss": 0.5809,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5451
    },
    {
      "epoch": 0.7413652434049497,
      "grad_norm": 0.4921875,
      "learning_rate": 3.80388856216188e-06,
      "loss": 0.6299,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5452
    },
    {
      "epoch": 0.7415012238237694,
      "grad_norm": 0.57421875,
      "learning_rate": 3.8001627015778355e-06,
      "loss": 0.5461,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5453
    },
    {
      "epoch": 0.741637204242589,
      "grad_norm": 0.357421875,
      "learning_rate": 3.7964382385129206e-06,
      "loss": 0.6929,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5454
    },
    {
      "epoch": 0.7417731846614087,
      "grad_norm": 0.326171875,
      "learning_rate": 3.7927151738066693e-06,
      "loss": 0.6032,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5455
    },
    {
      "epoch": 0.7419091650802284,
      "grad_norm": 0.318359375,
      "learning_rate": 3.788993508298311e-06,
      "loss": 0.6536,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5456
    },
    {
      "epoch": 0.7420451454990481,
      "grad_norm": 0.2451171875,
      "learning_rate": 3.785273242826749e-06,
      "loss": 0.4406,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5457
    },
    {
      "epoch": 0.7421811259178678,
      "grad_norm": 0.5,
      "learning_rate": 3.781554378230584e-06,
      "loss": 0.5356,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5458
    },
    {
      "epoch": 0.7423171063366876,
      "grad_norm": 0.375,
      "learning_rate": 3.777836915348081e-06,
      "loss": 0.7279,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5459
    },
    {
      "epoch": 0.7424530867555073,
      "grad_norm": 0.39453125,
      "learning_rate": 3.774120855017218e-06,
      "loss": 0.6598,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5460
    },
    {
      "epoch": 0.7425890671743269,
      "grad_norm": 1.0859375,
      "learning_rate": 3.7704061980756292e-06,
      "loss": 0.7542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5461
    },
    {
      "epoch": 0.7427250475931466,
      "grad_norm": 0.5,
      "learning_rate": 3.7666929453606493e-06,
      "loss": 0.8577,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5462
    },
    {
      "epoch": 0.7428610280119663,
      "grad_norm": 0.3046875,
      "learning_rate": 3.7629810977092852e-06,
      "loss": 0.6107,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5463
    },
    {
      "epoch": 0.742997008430786,
      "grad_norm": 0.32421875,
      "learning_rate": 3.7592706559582404e-06,
      "loss": 0.4645,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5464
    },
    {
      "epoch": 0.7431329888496057,
      "grad_norm": 0.259765625,
      "learning_rate": 3.755561620943885e-06,
      "loss": 0.584,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5465
    },
    {
      "epoch": 0.7432689692684253,
      "grad_norm": 0.41015625,
      "learning_rate": 3.751853993502288e-06,
      "loss": 0.6927,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5466
    },
    {
      "epoch": 0.743404949687245,
      "grad_norm": 0.42578125,
      "learning_rate": 3.7481477744691885e-06,
      "loss": 0.681,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5467
    },
    {
      "epoch": 0.7435409301060647,
      "grad_norm": 0.388671875,
      "learning_rate": 3.7444429646800206e-06,
      "loss": 0.7829,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5468
    },
    {
      "epoch": 0.7436769105248844,
      "grad_norm": 0.494140625,
      "learning_rate": 3.7407395649698853e-06,
      "loss": 0.7357,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5469
    },
    {
      "epoch": 0.7438128909437041,
      "grad_norm": 0.68359375,
      "learning_rate": 3.737037576173581e-06,
      "loss": 0.7357,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5470
    },
    {
      "epoch": 0.7439488713625237,
      "grad_norm": 0.4296875,
      "learning_rate": 3.7333369991255764e-06,
      "loss": 0.5083,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5471
    },
    {
      "epoch": 0.7440848517813435,
      "grad_norm": 0.341796875,
      "learning_rate": 3.7296378346600314e-06,
      "loss": 0.6305,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5472
    },
    {
      "epoch": 0.7442208322001632,
      "grad_norm": 0.376953125,
      "learning_rate": 3.7259400836107773e-06,
      "loss": 0.7152,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5473
    },
    {
      "epoch": 0.7443568126189829,
      "grad_norm": 0.7734375,
      "learning_rate": 3.7222437468113383e-06,
      "loss": 0.5312,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5474
    },
    {
      "epoch": 0.7444927930378026,
      "grad_norm": 0.38671875,
      "learning_rate": 3.718548825094909e-06,
      "loss": 0.6437,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5475
    },
    {
      "epoch": 0.7446287734566223,
      "grad_norm": 0.36328125,
      "learning_rate": 3.7148553192943736e-06,
      "loss": 0.6021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5476
    },
    {
      "epoch": 0.7447647538754419,
      "grad_norm": 0.64453125,
      "learning_rate": 3.7111632302422908e-06,
      "loss": 0.707,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5477
    },
    {
      "epoch": 0.7449007342942616,
      "grad_norm": 0.482421875,
      "learning_rate": 3.707472558770905e-06,
      "loss": 0.7726,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5478
    },
    {
      "epoch": 0.7450367147130813,
      "grad_norm": 0.40234375,
      "learning_rate": 3.7037833057121352e-06,
      "loss": 0.7607,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5479
    },
    {
      "epoch": 0.745172695131901,
      "grad_norm": 0.384765625,
      "learning_rate": 3.700095471897589e-06,
      "loss": 0.5781,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5480
    },
    {
      "epoch": 0.7453086755507207,
      "grad_norm": 2.46875,
      "learning_rate": 3.696409058158544e-06,
      "loss": 0.5653,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5481
    },
    {
      "epoch": 0.7454446559695403,
      "grad_norm": 0.373046875,
      "learning_rate": 3.692724065325969e-06,
      "loss": 0.6263,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5482
    },
    {
      "epoch": 0.74558063638836,
      "grad_norm": 0.353515625,
      "learning_rate": 3.689040494230499e-06,
      "loss": 0.7371,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5483
    },
    {
      "epoch": 0.7457166168071797,
      "grad_norm": 0.314453125,
      "learning_rate": 3.6853583457024634e-06,
      "loss": 0.6194,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5484
    },
    {
      "epoch": 0.7458525972259995,
      "grad_norm": 0.31640625,
      "learning_rate": 3.6816776205718565e-06,
      "loss": 0.6673,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5485
    },
    {
      "epoch": 0.7459885776448192,
      "grad_norm": 0.4453125,
      "learning_rate": 3.6779983196683655e-06,
      "loss": 0.7835,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5486
    },
    {
      "epoch": 0.7461245580636389,
      "grad_norm": 0.47265625,
      "learning_rate": 3.6743204438213465e-06,
      "loss": 0.7171,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5487
    },
    {
      "epoch": 0.7462605384824585,
      "grad_norm": 0.36328125,
      "learning_rate": 3.670643993859835e-06,
      "loss": 0.6942,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5488
    },
    {
      "epoch": 0.7463965189012782,
      "grad_norm": 0.376953125,
      "learning_rate": 3.666968970612553e-06,
      "loss": 0.6797,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5489
    },
    {
      "epoch": 0.7465324993200979,
      "grad_norm": 0.29296875,
      "learning_rate": 3.6632953749078902e-06,
      "loss": 0.5819,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5490
    },
    {
      "epoch": 0.7466684797389176,
      "grad_norm": 0.396484375,
      "learning_rate": 3.6596232075739267e-06,
      "loss": 0.3796,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5491
    },
    {
      "epoch": 0.7468044601577373,
      "grad_norm": 0.32421875,
      "learning_rate": 3.655952469438406e-06,
      "loss": 0.5365,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5492
    },
    {
      "epoch": 0.746940440576557,
      "grad_norm": 0.375,
      "learning_rate": 3.6522831613287656e-06,
      "loss": 0.6333,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5493
    },
    {
      "epoch": 0.7470764209953766,
      "grad_norm": 0.400390625,
      "learning_rate": 3.6486152840721046e-06,
      "loss": 0.7819,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5494
    },
    {
      "epoch": 0.7472124014141963,
      "grad_norm": 0.208984375,
      "learning_rate": 3.644948838495215e-06,
      "loss": 0.3569,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5495
    },
    {
      "epoch": 0.747348381833016,
      "grad_norm": 0.451171875,
      "learning_rate": 3.6412838254245497e-06,
      "loss": 0.7992,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5496
    },
    {
      "epoch": 0.7474843622518358,
      "grad_norm": 0.353515625,
      "learning_rate": 3.6376202456862574e-06,
      "loss": 0.716,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5497
    },
    {
      "epoch": 0.7476203426706555,
      "grad_norm": 0.328125,
      "learning_rate": 3.633958100106144e-06,
      "loss": 0.4699,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5498
    },
    {
      "epoch": 0.7477563230894752,
      "grad_norm": 0.267578125,
      "learning_rate": 3.6302973895097118e-06,
      "loss": 0.5552,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5499
    },
    {
      "epoch": 0.7478923035082948,
      "grad_norm": 0.357421875,
      "learning_rate": 3.62663811472212e-06,
      "loss": 0.7726,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5500
    },
    {
      "epoch": 0.7480282839271145,
      "grad_norm": 0.44140625,
      "learning_rate": 3.622980276568221e-06,
      "loss": 0.5396,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5501
    },
    {
      "epoch": 0.7481642643459342,
      "grad_norm": 0.376953125,
      "learning_rate": 3.6193238758725313e-06,
      "loss": 0.6915,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5502
    },
    {
      "epoch": 0.7483002447647539,
      "grad_norm": 0.4609375,
      "learning_rate": 3.6156689134592525e-06,
      "loss": 0.6755,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5503
    },
    {
      "epoch": 0.7484362251835736,
      "grad_norm": 0.38671875,
      "learning_rate": 3.6120153901522546e-06,
      "loss": 0.6652,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5504
    },
    {
      "epoch": 0.7485722056023932,
      "grad_norm": 0.3515625,
      "learning_rate": 3.60836330677509e-06,
      "loss": 0.5819,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5505
    },
    {
      "epoch": 0.7487081860212129,
      "grad_norm": 0.37890625,
      "learning_rate": 3.604712664150978e-06,
      "loss": 0.8797,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5506
    },
    {
      "epoch": 0.7488441664400326,
      "grad_norm": 0.33984375,
      "learning_rate": 3.601063463102823e-06,
      "loss": 0.6618,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5507
    },
    {
      "epoch": 0.7489801468588523,
      "grad_norm": 0.421875,
      "learning_rate": 3.5974157044531955e-06,
      "loss": 0.6213,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5508
    },
    {
      "epoch": 0.749116127277672,
      "grad_norm": 0.91015625,
      "learning_rate": 3.5937693890243484e-06,
      "loss": 0.9089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5509
    },
    {
      "epoch": 0.7492521076964918,
      "grad_norm": 0.33203125,
      "learning_rate": 3.5901245176382017e-06,
      "loss": 0.6948,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5510
    },
    {
      "epoch": 0.7493880881153114,
      "grad_norm": 0.287109375,
      "learning_rate": 3.5864810911163595e-06,
      "loss": 0.6688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5511
    },
    {
      "epoch": 0.7495240685341311,
      "grad_norm": 0.515625,
      "learning_rate": 3.5828391102800884e-06,
      "loss": 0.7594,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5512
    },
    {
      "epoch": 0.7496600489529508,
      "grad_norm": 0.435546875,
      "learning_rate": 3.579198575950342e-06,
      "loss": 0.6461,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5513
    },
    {
      "epoch": 0.7497960293717705,
      "grad_norm": 0.306640625,
      "learning_rate": 3.5755594889477343e-06,
      "loss": 0.4653,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5514
    },
    {
      "epoch": 0.7499320097905902,
      "grad_norm": 0.361328125,
      "learning_rate": 3.5719218500925666e-06,
      "loss": 0.5422,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5515
    },
    {
      "epoch": 0.7500679902094098,
      "grad_norm": 0.447265625,
      "learning_rate": 3.568285660204801e-06,
      "loss": 0.5425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5516
    },
    {
      "epoch": 0.7502039706282295,
      "grad_norm": 0.314453125,
      "learning_rate": 3.5646509201040857e-06,
      "loss": 0.5446,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5517
    },
    {
      "epoch": 0.7503399510470492,
      "grad_norm": 0.54296875,
      "learning_rate": 3.5610176306097287e-06,
      "loss": 0.7201,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5518
    },
    {
      "epoch": 0.7504759314658689,
      "grad_norm": 0.298828125,
      "learning_rate": 3.557385792540722e-06,
      "loss": 0.5544,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5519
    },
    {
      "epoch": 0.7506119118846886,
      "grad_norm": 0.51171875,
      "learning_rate": 3.553755406715724e-06,
      "loss": 0.6252,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5520
    },
    {
      "epoch": 0.7507478923035082,
      "grad_norm": 0.90625,
      "learning_rate": 3.5501264739530737e-06,
      "loss": 0.7417,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5521
    },
    {
      "epoch": 0.750883872722328,
      "grad_norm": 0.41796875,
      "learning_rate": 3.5464989950707697e-06,
      "loss": 0.7619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5522
    },
    {
      "epoch": 0.7510198531411477,
      "grad_norm": 0.4765625,
      "learning_rate": 3.542872970886496e-06,
      "loss": 0.8555,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5523
    },
    {
      "epoch": 0.7511558335599674,
      "grad_norm": 0.388671875,
      "learning_rate": 3.5392484022175976e-06,
      "loss": 0.7041,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5524
    },
    {
      "epoch": 0.7512918139787871,
      "grad_norm": 0.373046875,
      "learning_rate": 3.5356252898811017e-06,
      "loss": 0.6478,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5525
    },
    {
      "epoch": 0.7514277943976068,
      "grad_norm": 0.408203125,
      "learning_rate": 3.532003634693697e-06,
      "loss": 0.639,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5526
    },
    {
      "epoch": 0.7515637748164264,
      "grad_norm": 0.412109375,
      "learning_rate": 3.528383437471753e-06,
      "loss": 0.8182,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5527
    },
    {
      "epoch": 0.7516997552352461,
      "grad_norm": 0.5234375,
      "learning_rate": 3.524764699031302e-06,
      "loss": 0.7868,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5528
    },
    {
      "epoch": 0.7518357356540658,
      "grad_norm": 0.466796875,
      "learning_rate": 3.5211474201880592e-06,
      "loss": 0.7897,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5529
    },
    {
      "epoch": 0.7519717160728855,
      "grad_norm": 0.47265625,
      "learning_rate": 3.5175316017573935e-06,
      "loss": 0.8553,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5530
    },
    {
      "epoch": 0.7521076964917052,
      "grad_norm": 0.3359375,
      "learning_rate": 3.5139172445543647e-06,
      "loss": 0.6426,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5531
    },
    {
      "epoch": 0.7522436769105248,
      "grad_norm": 0.38671875,
      "learning_rate": 3.510304349393685e-06,
      "loss": 0.6909,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5532
    },
    {
      "epoch": 0.7523796573293445,
      "grad_norm": 0.333984375,
      "learning_rate": 3.506692917089751e-06,
      "loss": 0.7199,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5533
    },
    {
      "epoch": 0.7525156377481642,
      "grad_norm": 0.408203125,
      "learning_rate": 3.5030829484566185e-06,
      "loss": 0.7902,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5534
    },
    {
      "epoch": 0.752651618166984,
      "grad_norm": 0.5078125,
      "learning_rate": 3.499474444308024e-06,
      "loss": 0.8104,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5535
    },
    {
      "epoch": 0.7527875985858037,
      "grad_norm": 0.384765625,
      "learning_rate": 3.4958674054573627e-06,
      "loss": 0.7474,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5536
    },
    {
      "epoch": 0.7529235790046234,
      "grad_norm": 0.453125,
      "learning_rate": 3.4922618327177114e-06,
      "loss": 0.7002,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5537
    },
    {
      "epoch": 0.753059559423443,
      "grad_norm": 0.373046875,
      "learning_rate": 3.488657726901804e-06,
      "loss": 0.5827,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5538
    },
    {
      "epoch": 0.7531955398422627,
      "grad_norm": 0.400390625,
      "learning_rate": 3.4850550888220557e-06,
      "loss": 0.8133,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5539
    },
    {
      "epoch": 0.7533315202610824,
      "grad_norm": 0.330078125,
      "learning_rate": 3.481453919290544e-06,
      "loss": 0.7036,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5540
    },
    {
      "epoch": 0.7534675006799021,
      "grad_norm": 0.318359375,
      "learning_rate": 3.4778542191190113e-06,
      "loss": 0.6055,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5541
    },
    {
      "epoch": 0.7536034810987218,
      "grad_norm": 0.5234375,
      "learning_rate": 3.474255989118881e-06,
      "loss": 0.7645,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5542
    },
    {
      "epoch": 0.7537394615175415,
      "grad_norm": 0.5,
      "learning_rate": 3.470659230101231e-06,
      "loss": 0.4549,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5543
    },
    {
      "epoch": 0.7538754419363611,
      "grad_norm": 0.37890625,
      "learning_rate": 3.4670639428768236e-06,
      "loss": 0.7559,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5544
    },
    {
      "epoch": 0.7540114223551808,
      "grad_norm": 0.474609375,
      "learning_rate": 3.463470128256071e-06,
      "loss": 0.8524,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5545
    },
    {
      "epoch": 0.7541474027740005,
      "grad_norm": 1.1328125,
      "learning_rate": 3.4598777870490717e-06,
      "loss": 0.8836,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5546
    },
    {
      "epoch": 0.7542833831928203,
      "grad_norm": 0.3203125,
      "learning_rate": 3.456286920065576e-06,
      "loss": 0.7334,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5547
    },
    {
      "epoch": 0.75441936361164,
      "grad_norm": 0.3359375,
      "learning_rate": 3.4526975281150156e-06,
      "loss": 0.5895,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5548
    },
    {
      "epoch": 0.7545553440304597,
      "grad_norm": 0.396484375,
      "learning_rate": 3.449109612006477e-06,
      "loss": 0.6916,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5549
    },
    {
      "epoch": 0.7546913244492793,
      "grad_norm": 0.392578125,
      "learning_rate": 3.445523172548727e-06,
      "loss": 0.7757,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5550
    },
    {
      "epoch": 0.754827304868099,
      "grad_norm": 0.365234375,
      "learning_rate": 3.441938210550185e-06,
      "loss": 0.6138,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5551
    },
    {
      "epoch": 0.7549632852869187,
      "grad_norm": 0.36328125,
      "learning_rate": 3.4383547268189545e-06,
      "loss": 0.7617,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5552
    },
    {
      "epoch": 0.7550992657057384,
      "grad_norm": 0.53515625,
      "learning_rate": 3.434772722162788e-06,
      "loss": 0.5605,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5553
    },
    {
      "epoch": 0.7552352461245581,
      "grad_norm": 0.3125,
      "learning_rate": 3.4311921973891196e-06,
      "loss": 0.5798,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5554
    },
    {
      "epoch": 0.7553712265433777,
      "grad_norm": 0.83984375,
      "learning_rate": 3.4276131533050384e-06,
      "loss": 0.7681,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5555
    },
    {
      "epoch": 0.7555072069621974,
      "grad_norm": 0.72265625,
      "learning_rate": 3.4240355907173093e-06,
      "loss": 0.6242,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5556
    },
    {
      "epoch": 0.7556431873810171,
      "grad_norm": 0.3359375,
      "learning_rate": 3.4204595104323536e-06,
      "loss": 0.5791,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5557
    },
    {
      "epoch": 0.7557791677998368,
      "grad_norm": 0.87890625,
      "learning_rate": 3.41688491325627e-06,
      "loss": 0.6359,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5558
    },
    {
      "epoch": 0.7559151482186565,
      "grad_norm": 0.29296875,
      "learning_rate": 3.4133117999948086e-06,
      "loss": 0.5604,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5559
    },
    {
      "epoch": 0.7560511286374763,
      "grad_norm": 0.4921875,
      "learning_rate": 3.4097401714533997e-06,
      "loss": 0.7552,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5560
    },
    {
      "epoch": 0.7561871090562959,
      "grad_norm": 0.5078125,
      "learning_rate": 3.406170028437127e-06,
      "loss": 0.6213,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5561
    },
    {
      "epoch": 0.7563230894751156,
      "grad_norm": 1.296875,
      "learning_rate": 3.402601371750749e-06,
      "loss": 0.8143,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5562
    },
    {
      "epoch": 0.7564590698939353,
      "grad_norm": 0.51171875,
      "learning_rate": 3.399034202198679e-06,
      "loss": 0.5695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5563
    },
    {
      "epoch": 0.756595050312755,
      "grad_norm": 0.28515625,
      "learning_rate": 3.3954685205850068e-06,
      "loss": 0.5363,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5564
    },
    {
      "epoch": 0.7567310307315747,
      "grad_norm": 0.376953125,
      "learning_rate": 3.3919043277134755e-06,
      "loss": 0.6007,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5565
    },
    {
      "epoch": 0.7568670111503943,
      "grad_norm": 0.33203125,
      "learning_rate": 3.3883416243875024e-06,
      "loss": 0.5163,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5566
    },
    {
      "epoch": 0.757002991569214,
      "grad_norm": 0.625,
      "learning_rate": 3.384780411410158e-06,
      "loss": 0.6991,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5567
    },
    {
      "epoch": 0.7571389719880337,
      "grad_norm": 0.498046875,
      "learning_rate": 3.3812206895841913e-06,
      "loss": 0.7601,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5568
    },
    {
      "epoch": 0.7572749524068534,
      "grad_norm": 0.32421875,
      "learning_rate": 3.377662459711999e-06,
      "loss": 0.4448,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5569
    },
    {
      "epoch": 0.7574109328256731,
      "grad_norm": 0.322265625,
      "learning_rate": 3.3741057225956575e-06,
      "loss": 0.5671,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5570
    },
    {
      "epoch": 0.7575469132444927,
      "grad_norm": 0.369140625,
      "learning_rate": 3.3705504790368904e-06,
      "loss": 0.4701,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5571
    },
    {
      "epoch": 0.7576828936633124,
      "grad_norm": 0.41015625,
      "learning_rate": 3.366996729837102e-06,
      "loss": 0.6362,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5572
    },
    {
      "epoch": 0.7578188740821322,
      "grad_norm": 0.421875,
      "learning_rate": 3.3634444757973417e-06,
      "loss": 0.769,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5573
    },
    {
      "epoch": 0.7579548545009519,
      "grad_norm": 0.318359375,
      "learning_rate": 3.35989371771834e-06,
      "loss": 0.5356,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5574
    },
    {
      "epoch": 0.7580908349197716,
      "grad_norm": 0.42578125,
      "learning_rate": 3.3563444564004732e-06,
      "loss": 0.8384,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5575
    },
    {
      "epoch": 0.7582268153385913,
      "grad_norm": 0.3203125,
      "learning_rate": 3.3527966926437948e-06,
      "loss": 0.5674,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5576
    },
    {
      "epoch": 0.758362795757411,
      "grad_norm": 0.578125,
      "learning_rate": 3.349250427248009e-06,
      "loss": 0.5675,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5577
    },
    {
      "epoch": 0.7584987761762306,
      "grad_norm": 0.55078125,
      "learning_rate": 3.3457056610124905e-06,
      "loss": 0.7074,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5578
    },
    {
      "epoch": 0.7586347565950503,
      "grad_norm": 0.498046875,
      "learning_rate": 3.342162394736267e-06,
      "loss": 0.5355,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5579
    },
    {
      "epoch": 0.75877073701387,
      "grad_norm": 0.373046875,
      "learning_rate": 3.3386206292180456e-06,
      "loss": 0.638,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5580
    },
    {
      "epoch": 0.7589067174326897,
      "grad_norm": 0.345703125,
      "learning_rate": 3.3350803652561736e-06,
      "loss": 0.589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5581
    },
    {
      "epoch": 0.7590426978515094,
      "grad_norm": 0.330078125,
      "learning_rate": 3.3315416036486756e-06,
      "loss": 0.5972,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5582
    },
    {
      "epoch": 0.759178678270329,
      "grad_norm": 0.376953125,
      "learning_rate": 3.3280043451932266e-06,
      "loss": 0.7703,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5583
    },
    {
      "epoch": 0.7593146586891487,
      "grad_norm": 0.46875,
      "learning_rate": 3.324468590687173e-06,
      "loss": 0.4474,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5584
    },
    {
      "epoch": 0.7594506391079685,
      "grad_norm": 0.376953125,
      "learning_rate": 3.320934340927513e-06,
      "loss": 0.6444,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5585
    },
    {
      "epoch": 0.7595866195267882,
      "grad_norm": 0.33984375,
      "learning_rate": 3.317401596710913e-06,
      "loss": 0.556,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5586
    },
    {
      "epoch": 0.7597225999456079,
      "grad_norm": 0.58984375,
      "learning_rate": 3.313870358833693e-06,
      "loss": 0.735,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5587
    },
    {
      "epoch": 0.7598585803644275,
      "grad_norm": 0.55078125,
      "learning_rate": 3.3103406280918415e-06,
      "loss": 0.6017,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5588
    },
    {
      "epoch": 0.7599945607832472,
      "grad_norm": 0.373046875,
      "learning_rate": 3.306812405280998e-06,
      "loss": 0.645,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5589
    },
    {
      "epoch": 0.7601305412020669,
      "grad_norm": 0.421875,
      "learning_rate": 3.3032856911964728e-06,
      "loss": 0.5542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5590
    },
    {
      "epoch": 0.7602665216208866,
      "grad_norm": 0.4375,
      "learning_rate": 3.2997604866332246e-06,
      "loss": 0.783,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5591
    },
    {
      "epoch": 0.7604025020397063,
      "grad_norm": 0.3203125,
      "learning_rate": 3.296236792385884e-06,
      "loss": 0.4398,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5592
    },
    {
      "epoch": 0.760538482458526,
      "grad_norm": 0.51953125,
      "learning_rate": 3.292714609248727e-06,
      "loss": 0.7453,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5593
    },
    {
      "epoch": 0.7606744628773456,
      "grad_norm": 0.392578125,
      "learning_rate": 3.2891939380157044e-06,
      "loss": 0.599,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5594
    },
    {
      "epoch": 0.7608104432961653,
      "grad_norm": 0.28515625,
      "learning_rate": 3.2856747794804145e-06,
      "loss": 0.5477,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5595
    },
    {
      "epoch": 0.760946423714985,
      "grad_norm": 0.34375,
      "learning_rate": 3.2821571344361168e-06,
      "loss": 0.5496,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5596
    },
    {
      "epoch": 0.7610824041338047,
      "grad_norm": 0.37890625,
      "learning_rate": 3.2786410036757366e-06,
      "loss": 0.7559,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5597
    },
    {
      "epoch": 0.7612183845526245,
      "grad_norm": 0.37109375,
      "learning_rate": 3.275126387991847e-06,
      "loss": 0.6372,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5598
    },
    {
      "epoch": 0.7613543649714442,
      "grad_norm": 0.86328125,
      "learning_rate": 3.2716132881766916e-06,
      "loss": 0.7466,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5599
    },
    {
      "epoch": 0.7614903453902638,
      "grad_norm": 0.283203125,
      "learning_rate": 3.26810170502216e-06,
      "loss": 0.501,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5600
    },
    {
      "epoch": 0.7616263258090835,
      "grad_norm": 0.5234375,
      "learning_rate": 3.2645916393198107e-06,
      "loss": 0.77,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5601
    },
    {
      "epoch": 0.7617623062279032,
      "grad_norm": 0.330078125,
      "learning_rate": 3.2610830918608516e-06,
      "loss": 0.7095,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5602
    },
    {
      "epoch": 0.7618982866467229,
      "grad_norm": 0.283203125,
      "learning_rate": 3.2575760634361576e-06,
      "loss": 0.472,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5603
    },
    {
      "epoch": 0.7620342670655426,
      "grad_norm": 0.251953125,
      "learning_rate": 3.2540705548362485e-06,
      "loss": 0.4849,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5604
    },
    {
      "epoch": 0.7621702474843622,
      "grad_norm": 0.28515625,
      "learning_rate": 3.2505665668513154e-06,
      "loss": 0.5511,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5605
    },
    {
      "epoch": 0.7623062279031819,
      "grad_norm": 0.6484375,
      "learning_rate": 3.2470641002711945e-06,
      "loss": 0.7886,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5606
    },
    {
      "epoch": 0.7624422083220016,
      "grad_norm": 0.5234375,
      "learning_rate": 3.243563155885391e-06,
      "loss": 0.7405,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5607
    },
    {
      "epoch": 0.7625781887408213,
      "grad_norm": 0.67578125,
      "learning_rate": 3.240063734483052e-06,
      "loss": 0.6784,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5608
    },
    {
      "epoch": 0.762714169159641,
      "grad_norm": 0.326171875,
      "learning_rate": 3.2365658368529997e-06,
      "loss": 0.621,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5609
    },
    {
      "epoch": 0.7628501495784608,
      "grad_norm": 0.408203125,
      "learning_rate": 3.2330694637836947e-06,
      "loss": 0.5835,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5610
    },
    {
      "epoch": 0.7629861299972804,
      "grad_norm": 0.4765625,
      "learning_rate": 3.229574616063268e-06,
      "loss": 0.7427,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5611
    },
    {
      "epoch": 0.7631221104161001,
      "grad_norm": 0.2392578125,
      "learning_rate": 3.226081294479496e-06,
      "loss": 0.4502,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5612
    },
    {
      "epoch": 0.7632580908349198,
      "grad_norm": 0.49609375,
      "learning_rate": 3.2225894998198216e-06,
      "loss": 0.6979,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5613
    },
    {
      "epoch": 0.7633940712537395,
      "grad_norm": 0.412109375,
      "learning_rate": 3.219099232871332e-06,
      "loss": 0.799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5614
    },
    {
      "epoch": 0.7635300516725592,
      "grad_norm": 0.546875,
      "learning_rate": 3.215610494420782e-06,
      "loss": 0.6315,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5615
    },
    {
      "epoch": 0.7636660320913788,
      "grad_norm": 0.734375,
      "learning_rate": 3.2121232852545705e-06,
      "loss": 0.7673,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5616
    },
    {
      "epoch": 0.7638020125101985,
      "grad_norm": 0.44140625,
      "learning_rate": 3.2086376061587623e-06,
      "loss": 0.668,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5617
    },
    {
      "epoch": 0.7639379929290182,
      "grad_norm": 0.484375,
      "learning_rate": 3.205153457919067e-06,
      "loss": 0.8623,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5618
    },
    {
      "epoch": 0.7640739733478379,
      "grad_norm": 0.396484375,
      "learning_rate": 3.2016708413208584e-06,
      "loss": 0.7611,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5619
    },
    {
      "epoch": 0.7642099537666576,
      "grad_norm": 0.373046875,
      "learning_rate": 3.198189757149157e-06,
      "loss": 0.4425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5620
    },
    {
      "epoch": 0.7643459341854772,
      "grad_norm": 0.390625,
      "learning_rate": 3.194710206188647e-06,
      "loss": 0.6629,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5621
    },
    {
      "epoch": 0.7644819146042969,
      "grad_norm": 0.41796875,
      "learning_rate": 3.1912321892236555e-06,
      "loss": 0.721,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5622
    },
    {
      "epoch": 0.7646178950231167,
      "grad_norm": 0.5234375,
      "learning_rate": 3.1877557070381747e-06,
      "loss": 0.8664,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5623
    },
    {
      "epoch": 0.7647538754419364,
      "grad_norm": 0.47265625,
      "learning_rate": 3.184280760415843e-06,
      "loss": 0.757,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5624
    },
    {
      "epoch": 0.7648898558607561,
      "grad_norm": 0.359375,
      "learning_rate": 3.1808073501399606e-06,
      "loss": 0.7043,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5625
    },
    {
      "epoch": 0.7650258362795758,
      "grad_norm": 0.45703125,
      "learning_rate": 3.1773354769934693e-06,
      "loss": 0.6453,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5626
    },
    {
      "epoch": 0.7651618166983954,
      "grad_norm": 0.423828125,
      "learning_rate": 3.1738651417589795e-06,
      "loss": 0.6989,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5627
    },
    {
      "epoch": 0.7652977971172151,
      "grad_norm": 0.50390625,
      "learning_rate": 3.1703963452187415e-06,
      "loss": 0.6926,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5628
    },
    {
      "epoch": 0.7654337775360348,
      "grad_norm": 0.384765625,
      "learning_rate": 3.166929088154669e-06,
      "loss": 0.708,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5629
    },
    {
      "epoch": 0.7655697579548545,
      "grad_norm": 0.51953125,
      "learning_rate": 3.1634633713483174e-06,
      "loss": 0.4666,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5630
    },
    {
      "epoch": 0.7657057383736742,
      "grad_norm": 0.41796875,
      "learning_rate": 3.1599991955809104e-06,
      "loss": 0.8374,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5631
    },
    {
      "epoch": 0.7658417187924939,
      "grad_norm": 0.318359375,
      "learning_rate": 3.1565365616333076e-06,
      "loss": 0.5594,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5632
    },
    {
      "epoch": 0.7659776992113135,
      "grad_norm": 0.66015625,
      "learning_rate": 3.1530754702860355e-06,
      "loss": 0.7629,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5633
    },
    {
      "epoch": 0.7661136796301332,
      "grad_norm": 0.330078125,
      "learning_rate": 3.14961592231926e-06,
      "loss": 0.6076,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5634
    },
    {
      "epoch": 0.7662496600489529,
      "grad_norm": 0.5,
      "learning_rate": 3.1461579185128133e-06,
      "loss": 0.7892,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5635
    },
    {
      "epoch": 0.7663856404677727,
      "grad_norm": 0.427734375,
      "learning_rate": 3.1427014596461636e-06,
      "loss": 0.6882,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5636
    },
    {
      "epoch": 0.7665216208865924,
      "grad_norm": 0.40234375,
      "learning_rate": 3.1392465464984455e-06,
      "loss": 0.687,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5637
    },
    {
      "epoch": 0.766657601305412,
      "grad_norm": 0.3125,
      "learning_rate": 3.135793179848433e-06,
      "loss": 0.7096,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5638
    },
    {
      "epoch": 0.7667935817242317,
      "grad_norm": 0.388671875,
      "learning_rate": 3.13234136047456e-06,
      "loss": 0.5545,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5639
    },
    {
      "epoch": 0.7669295621430514,
      "grad_norm": 0.490234375,
      "learning_rate": 3.1288910891549096e-06,
      "loss": 0.8138,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5640
    },
    {
      "epoch": 0.7670655425618711,
      "grad_norm": 0.259765625,
      "learning_rate": 3.1254423666672175e-06,
      "loss": 0.4468,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5641
    },
    {
      "epoch": 0.7672015229806908,
      "grad_norm": 0.330078125,
      "learning_rate": 3.1219951937888606e-06,
      "loss": 0.6715,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5642
    },
    {
      "epoch": 0.7673375033995105,
      "grad_norm": 0.30859375,
      "learning_rate": 3.1185495712968827e-06,
      "loss": 0.6476,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5643
    },
    {
      "epoch": 0.7674734838183301,
      "grad_norm": 0.890625,
      "learning_rate": 3.11510549996796e-06,
      "loss": 0.6099,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5644
    },
    {
      "epoch": 0.7676094642371498,
      "grad_norm": 0.2490234375,
      "learning_rate": 3.1116629805784373e-06,
      "loss": 0.5335,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5645
    },
    {
      "epoch": 0.7677454446559695,
      "grad_norm": 0.58203125,
      "learning_rate": 3.108222013904292e-06,
      "loss": 0.7458,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5646
    },
    {
      "epoch": 0.7678814250747892,
      "grad_norm": 0.2412109375,
      "learning_rate": 3.104782600721168e-06,
      "loss": 0.4963,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5647
    },
    {
      "epoch": 0.768017405493609,
      "grad_norm": 0.37109375,
      "learning_rate": 3.101344741804346e-06,
      "loss": 0.6055,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5648
    },
    {
      "epoch": 0.7681533859124287,
      "grad_norm": 0.48828125,
      "learning_rate": 3.0979084379287615e-06,
      "loss": 0.7975,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5649
    },
    {
      "epoch": 0.7682893663312483,
      "grad_norm": 0.46875,
      "learning_rate": 3.094473689869002e-06,
      "loss": 0.7384,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5650
    },
    {
      "epoch": 0.768425346750068,
      "grad_norm": 0.6171875,
      "learning_rate": 3.091040498399298e-06,
      "loss": 0.8911,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5651
    },
    {
      "epoch": 0.7685613271688877,
      "grad_norm": 0.353515625,
      "learning_rate": 3.0876088642935387e-06,
      "loss": 0.7293,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5652
    },
    {
      "epoch": 0.7686973075877074,
      "grad_norm": 0.515625,
      "learning_rate": 3.0841787883252494e-06,
      "loss": 0.5881,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5653
    },
    {
      "epoch": 0.7688332880065271,
      "grad_norm": 0.400390625,
      "learning_rate": 3.0807502712676174e-06,
      "loss": 0.521,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5654
    },
    {
      "epoch": 0.7689692684253467,
      "grad_norm": 0.396484375,
      "learning_rate": 3.077323313893467e-06,
      "loss": 0.7279,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5655
    },
    {
      "epoch": 0.7691052488441664,
      "grad_norm": 0.62890625,
      "learning_rate": 3.0738979169752813e-06,
      "loss": 0.6927,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5656
    },
    {
      "epoch": 0.7692412292629861,
      "grad_norm": 0.53515625,
      "learning_rate": 3.0704740812851807e-06,
      "loss": 0.8006,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5657
    },
    {
      "epoch": 0.7693772096818058,
      "grad_norm": 0.6328125,
      "learning_rate": 3.0670518075949475e-06,
      "loss": 0.7456,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5658
    },
    {
      "epoch": 0.7695131901006255,
      "grad_norm": 0.66015625,
      "learning_rate": 3.063631096675996e-06,
      "loss": 0.7454,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5659
    },
    {
      "epoch": 0.7696491705194451,
      "grad_norm": 0.4765625,
      "learning_rate": 3.0602119492994033e-06,
      "loss": 0.8564,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5660
    },
    {
      "epoch": 0.7697851509382649,
      "grad_norm": 0.287109375,
      "learning_rate": 3.0567943662358803e-06,
      "loss": 0.5897,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5661
    },
    {
      "epoch": 0.7699211313570846,
      "grad_norm": 0.259765625,
      "learning_rate": 3.0533783482557977e-06,
      "loss": 0.5666,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5662
    },
    {
      "epoch": 0.7700571117759043,
      "grad_norm": 0.4921875,
      "learning_rate": 3.0499638961291623e-06,
      "loss": 0.6424,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5663
    },
    {
      "epoch": 0.770193092194724,
      "grad_norm": 0.58984375,
      "learning_rate": 3.0465510106256403e-06,
      "loss": 0.4779,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5664
    },
    {
      "epoch": 0.7703290726135437,
      "grad_norm": 0.64453125,
      "learning_rate": 3.0431396925145307e-06,
      "loss": 0.4636,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5665
    },
    {
      "epoch": 0.7704650530323633,
      "grad_norm": 0.6484375,
      "learning_rate": 3.039729942564793e-06,
      "loss": 0.5614,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5666
    },
    {
      "epoch": 0.770601033451183,
      "grad_norm": 0.58203125,
      "learning_rate": 3.0363217615450192e-06,
      "loss": 0.8421,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5667
    },
    {
      "epoch": 0.7707370138700027,
      "grad_norm": 0.337890625,
      "learning_rate": 3.032915150223462e-06,
      "loss": 0.7935,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5668
    },
    {
      "epoch": 0.7708729942888224,
      "grad_norm": 0.443359375,
      "learning_rate": 3.0295101093680067e-06,
      "loss": 0.8989,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5669
    },
    {
      "epoch": 0.7710089747076421,
      "grad_norm": 0.546875,
      "learning_rate": 3.026106639746199e-06,
      "loss": 0.6514,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5670
    },
    {
      "epoch": 0.7711449551264618,
      "grad_norm": 0.37890625,
      "learning_rate": 3.022704742125213e-06,
      "loss": 0.6491,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5671
    },
    {
      "epoch": 0.7712809355452814,
      "grad_norm": 0.37109375,
      "learning_rate": 3.0193044172718877e-06,
      "loss": 0.5848,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5672
    },
    {
      "epoch": 0.7714169159641012,
      "grad_norm": 0.359375,
      "learning_rate": 3.0159056659526896e-06,
      "loss": 0.4455,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5673
    },
    {
      "epoch": 0.7715528963829209,
      "grad_norm": 0.29296875,
      "learning_rate": 3.0125084889337465e-06,
      "loss": 0.5575,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5674
    },
    {
      "epoch": 0.7716888768017406,
      "grad_norm": 0.314453125,
      "learning_rate": 3.009112886980815e-06,
      "loss": 0.5791,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5675
    },
    {
      "epoch": 0.7718248572205603,
      "grad_norm": 0.6640625,
      "learning_rate": 3.0057188608593146e-06,
      "loss": 0.5031,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5676
    },
    {
      "epoch": 0.77196083763938,
      "grad_norm": 0.58984375,
      "learning_rate": 3.002326411334291e-06,
      "loss": 0.7231,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5677
    },
    {
      "epoch": 0.7720968180581996,
      "grad_norm": 0.298828125,
      "learning_rate": 2.9989355391704523e-06,
      "loss": 0.4603,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5678
    },
    {
      "epoch": 0.7722327984770193,
      "grad_norm": 0.62890625,
      "learning_rate": 2.995546245132136e-06,
      "loss": 0.7827,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5679
    },
    {
      "epoch": 0.772368778895839,
      "grad_norm": 0.36328125,
      "learning_rate": 2.9921585299833355e-06,
      "loss": 0.7913,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5680
    },
    {
      "epoch": 0.7725047593146587,
      "grad_norm": 0.4296875,
      "learning_rate": 2.9887723944876768e-06,
      "loss": 0.8836,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5681
    },
    {
      "epoch": 0.7726407397334784,
      "grad_norm": 0.33203125,
      "learning_rate": 2.9853878394084447e-06,
      "loss": 0.5866,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5682
    },
    {
      "epoch": 0.772776720152298,
      "grad_norm": 0.390625,
      "learning_rate": 2.982004865508551e-06,
      "loss": 0.7806,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5683
    },
    {
      "epoch": 0.7729127005711177,
      "grad_norm": 1.109375,
      "learning_rate": 2.978623473550567e-06,
      "loss": 0.7609,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5684
    },
    {
      "epoch": 0.7730486809899374,
      "grad_norm": 0.62109375,
      "learning_rate": 2.975243664296692e-06,
      "loss": 0.7811,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5685
    },
    {
      "epoch": 0.7731846614087572,
      "grad_norm": 0.478515625,
      "learning_rate": 2.971865438508784e-06,
      "loss": 0.7495,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5686
    },
    {
      "epoch": 0.7733206418275769,
      "grad_norm": 0.28125,
      "learning_rate": 2.968488796948329e-06,
      "loss": 0.5267,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5687
    },
    {
      "epoch": 0.7734566222463966,
      "grad_norm": 0.51171875,
      "learning_rate": 2.9651137403764706e-06,
      "loss": 0.4517,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5688
    },
    {
      "epoch": 0.7735926026652162,
      "grad_norm": 0.4296875,
      "learning_rate": 2.961740269553981e-06,
      "loss": 0.5954,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5689
    },
    {
      "epoch": 0.7737285830840359,
      "grad_norm": 0.375,
      "learning_rate": 2.9583683852412882e-06,
      "loss": 0.7729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5690
    },
    {
      "epoch": 0.7738645635028556,
      "grad_norm": 0.5,
      "learning_rate": 2.9549980881984496e-06,
      "loss": 0.4966,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5691
    },
    {
      "epoch": 0.7740005439216753,
      "grad_norm": 0.7109375,
      "learning_rate": 2.951629379185178e-06,
      "loss": 0.7479,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5692
    },
    {
      "epoch": 0.774136524340495,
      "grad_norm": 0.341796875,
      "learning_rate": 2.9482622589608143e-06,
      "loss": 0.5169,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5693
    },
    {
      "epoch": 0.7742725047593146,
      "grad_norm": 0.46484375,
      "learning_rate": 2.944896728284358e-06,
      "loss": 0.7803,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5694
    },
    {
      "epoch": 0.7744084851781343,
      "grad_norm": 0.37890625,
      "learning_rate": 2.9415327879144294e-06,
      "loss": 0.5474,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5695
    },
    {
      "epoch": 0.774544465596954,
      "grad_norm": 0.59765625,
      "learning_rate": 2.9381704386093125e-06,
      "loss": 0.5592,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5696
    },
    {
      "epoch": 0.7746804460157737,
      "grad_norm": 0.3515625,
      "learning_rate": 2.934809681126913e-06,
      "loss": 0.7074,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5697
    },
    {
      "epoch": 0.7748164264345934,
      "grad_norm": 0.33203125,
      "learning_rate": 2.9314505162247952e-06,
      "loss": 0.6159,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5698
    },
    {
      "epoch": 0.7749524068534132,
      "grad_norm": 0.5859375,
      "learning_rate": 2.928092944660146e-06,
      "loss": 0.6733,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5699
    },
    {
      "epoch": 0.7750883872722328,
      "grad_norm": 0.62890625,
      "learning_rate": 2.9247369671898142e-06,
      "loss": 0.6551,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5700
    },
    {
      "epoch": 0.7752243676910525,
      "grad_norm": 0.353515625,
      "learning_rate": 2.921382584570269e-06,
      "loss": 0.6885,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5701
    },
    {
      "epoch": 0.7753603481098722,
      "grad_norm": 0.388671875,
      "learning_rate": 2.9180297975576368e-06,
      "loss": 0.6873,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5702
    },
    {
      "epoch": 0.7754963285286919,
      "grad_norm": 0.458984375,
      "learning_rate": 2.914678606907673e-06,
      "loss": 0.6178,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5703
    },
    {
      "epoch": 0.7756323089475116,
      "grad_norm": 0.365234375,
      "learning_rate": 2.911329013375773e-06,
      "loss": 0.6527,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5704
    },
    {
      "epoch": 0.7757682893663312,
      "grad_norm": 0.79296875,
      "learning_rate": 2.9079810177169833e-06,
      "loss": 0.887,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5705
    },
    {
      "epoch": 0.7759042697851509,
      "grad_norm": 0.30859375,
      "learning_rate": 2.9046346206859776e-06,
      "loss": 0.6868,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5706
    },
    {
      "epoch": 0.7760402502039706,
      "grad_norm": 0.65625,
      "learning_rate": 2.901289823037079e-06,
      "loss": 0.6898,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5707
    },
    {
      "epoch": 0.7761762306227903,
      "grad_norm": 0.2353515625,
      "learning_rate": 2.8979466255242417e-06,
      "loss": 0.4202,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5708
    },
    {
      "epoch": 0.77631221104161,
      "grad_norm": 0.78125,
      "learning_rate": 2.8946050289010675e-06,
      "loss": 0.6745,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5709
    },
    {
      "epoch": 0.7764481914604296,
      "grad_norm": 0.37890625,
      "learning_rate": 2.891265033920788e-06,
      "loss": 0.6532,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5710
    },
    {
      "epoch": 0.7765841718792494,
      "grad_norm": 0.40625,
      "learning_rate": 2.887926641336285e-06,
      "loss": 0.7323,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5711
    },
    {
      "epoch": 0.7767201522980691,
      "grad_norm": 1.125,
      "learning_rate": 2.884589851900067e-06,
      "loss": 0.9619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5712
    },
    {
      "epoch": 0.7768561327168888,
      "grad_norm": 0.49609375,
      "learning_rate": 2.8812546663642927e-06,
      "loss": 0.7539,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5713
    },
    {
      "epoch": 0.7769921131357085,
      "grad_norm": 0.59375,
      "learning_rate": 2.8779210854807473e-06,
      "loss": 0.7052,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5714
    },
    {
      "epoch": 0.7771280935545282,
      "grad_norm": 0.7734375,
      "learning_rate": 2.8745891100008683e-06,
      "loss": 0.8506,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5715
    },
    {
      "epoch": 0.7772640739733478,
      "grad_norm": 0.421875,
      "learning_rate": 2.8712587406757165e-06,
      "loss": 0.5358,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5716
    },
    {
      "epoch": 0.7774000543921675,
      "grad_norm": 0.2294921875,
      "learning_rate": 2.8679299782560045e-06,
      "loss": 0.4092,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5717
    },
    {
      "epoch": 0.7775360348109872,
      "grad_norm": 0.40234375,
      "learning_rate": 2.864602823492069e-06,
      "loss": 0.6388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5718
    },
    {
      "epoch": 0.7776720152298069,
      "grad_norm": 0.380859375,
      "learning_rate": 2.861277277133898e-06,
      "loss": 0.6688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5719
    },
    {
      "epoch": 0.7778079956486266,
      "grad_norm": 0.33984375,
      "learning_rate": 2.857953339931104e-06,
      "loss": 0.5885,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5720
    },
    {
      "epoch": 0.7779439760674463,
      "grad_norm": 0.369140625,
      "learning_rate": 2.85463101263295e-06,
      "loss": 0.6547,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5721
    },
    {
      "epoch": 0.7780799564862659,
      "grad_norm": 0.38671875,
      "learning_rate": 2.851310295988323e-06,
      "loss": 0.6851,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5722
    },
    {
      "epoch": 0.7782159369050856,
      "grad_norm": 0.59765625,
      "learning_rate": 2.847991190745758e-06,
      "loss": 0.7274,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5723
    },
    {
      "epoch": 0.7783519173239054,
      "grad_norm": 0.30859375,
      "learning_rate": 2.844673697653417e-06,
      "loss": 0.5384,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5724
    },
    {
      "epoch": 0.7784878977427251,
      "grad_norm": 0.333984375,
      "learning_rate": 2.8413578174591094e-06,
      "loss": 0.6493,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5725
    },
    {
      "epoch": 0.7786238781615448,
      "grad_norm": 0.482421875,
      "learning_rate": 2.838043550910269e-06,
      "loss": 0.693,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5726
    },
    {
      "epoch": 0.7787598585803645,
      "grad_norm": 0.341796875,
      "learning_rate": 2.834730898753977e-06,
      "loss": 0.6326,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5727
    },
    {
      "epoch": 0.7788958389991841,
      "grad_norm": 0.375,
      "learning_rate": 2.8314198617369403e-06,
      "loss": 0.7244,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5728
    },
    {
      "epoch": 0.7790318194180038,
      "grad_norm": 0.28515625,
      "learning_rate": 2.828110440605515e-06,
      "loss": 0.5659,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5729
    },
    {
      "epoch": 0.7791677998368235,
      "grad_norm": 0.34765625,
      "learning_rate": 2.8248026361056757e-06,
      "loss": 0.5742,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5730
    },
    {
      "epoch": 0.7793037802556432,
      "grad_norm": 0.298828125,
      "learning_rate": 2.821496448983051e-06,
      "loss": 0.5609,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5731
    },
    {
      "epoch": 0.7794397606744629,
      "grad_norm": 0.330078125,
      "learning_rate": 2.818191879982888e-06,
      "loss": 0.7085,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5732
    },
    {
      "epoch": 0.7795757410932825,
      "grad_norm": 0.62109375,
      "learning_rate": 2.814888929850084e-06,
      "loss": 0.9023,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5733
    },
    {
      "epoch": 0.7797117215121022,
      "grad_norm": 0.5234375,
      "learning_rate": 2.8115875993291573e-06,
      "loss": 0.6768,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5734
    },
    {
      "epoch": 0.7798477019309219,
      "grad_norm": 0.3359375,
      "learning_rate": 2.8082878891642764e-06,
      "loss": 0.5812,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5735
    },
    {
      "epoch": 0.7799836823497417,
      "grad_norm": 0.396484375,
      "learning_rate": 2.804989800099227e-06,
      "loss": 0.5773,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5736
    },
    {
      "epoch": 0.7801196627685614,
      "grad_norm": 0.2578125,
      "learning_rate": 2.8016933328774475e-06,
      "loss": 0.5871,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5737
    },
    {
      "epoch": 0.780255643187381,
      "grad_norm": 0.39453125,
      "learning_rate": 2.7983984882419945e-06,
      "loss": 0.7782,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5738
    },
    {
      "epoch": 0.7803916236062007,
      "grad_norm": 0.5234375,
      "learning_rate": 2.795105266935573e-06,
      "loss": 0.6737,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5739
    },
    {
      "epoch": 0.7805276040250204,
      "grad_norm": 0.27734375,
      "learning_rate": 2.791813669700508e-06,
      "loss": 0.4237,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5740
    },
    {
      "epoch": 0.7806635844438401,
      "grad_norm": 0.453125,
      "learning_rate": 2.7885236972787733e-06,
      "loss": 0.8577,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5741
    },
    {
      "epoch": 0.7807995648626598,
      "grad_norm": 0.5390625,
      "learning_rate": 2.7852353504119622e-06,
      "loss": 0.8326,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5742
    },
    {
      "epoch": 0.7809355452814795,
      "grad_norm": 0.357421875,
      "learning_rate": 2.7819486298413147e-06,
      "loss": 0.5654,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5743
    },
    {
      "epoch": 0.7810715257002991,
      "grad_norm": 0.294921875,
      "learning_rate": 2.7786635363076907e-06,
      "loss": 0.4793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5744
    },
    {
      "epoch": 0.7812075061191188,
      "grad_norm": 0.373046875,
      "learning_rate": 2.7753800705515977e-06,
      "loss": 0.5404,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5745
    },
    {
      "epoch": 0.7813434865379385,
      "grad_norm": 0.373046875,
      "learning_rate": 2.772098233313163e-06,
      "loss": 0.4742,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5746
    },
    {
      "epoch": 0.7814794669567582,
      "grad_norm": 0.74609375,
      "learning_rate": 2.768818025332158e-06,
      "loss": 0.6803,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5747
    },
    {
      "epoch": 0.7816154473755779,
      "grad_norm": 0.6171875,
      "learning_rate": 2.7655394473479756e-06,
      "loss": 0.887,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5748
    },
    {
      "epoch": 0.7817514277943977,
      "grad_norm": 0.41015625,
      "learning_rate": 2.762262500099655e-06,
      "loss": 0.5636,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5749
    },
    {
      "epoch": 0.7818874082132173,
      "grad_norm": 0.375,
      "learning_rate": 2.758987184325852e-06,
      "loss": 0.6116,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5750
    },
    {
      "epoch": 0.782023388632037,
      "grad_norm": 0.33203125,
      "learning_rate": 2.7557135007648707e-06,
      "loss": 0.5794,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5751
    },
    {
      "epoch": 0.7821593690508567,
      "grad_norm": 0.486328125,
      "learning_rate": 2.7524414501546325e-06,
      "loss": 0.8675,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5752
    },
    {
      "epoch": 0.7822953494696764,
      "grad_norm": 0.421875,
      "learning_rate": 2.7491710332327025e-06,
      "loss": 0.7399,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5753
    },
    {
      "epoch": 0.7824313298884961,
      "grad_norm": 0.35546875,
      "learning_rate": 2.7459022507362687e-06,
      "loss": 0.6979,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5754
    },
    {
      "epoch": 0.7825673103073157,
      "grad_norm": 0.671875,
      "learning_rate": 2.7426351034021604e-06,
      "loss": 0.7197,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5755
    },
    {
      "epoch": 0.7827032907261354,
      "grad_norm": 0.26171875,
      "learning_rate": 2.739369591966825e-06,
      "loss": 0.5097,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5756
    },
    {
      "epoch": 0.7828392711449551,
      "grad_norm": 0.41796875,
      "learning_rate": 2.7361057171663554e-06,
      "loss": 0.6567,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5757
    },
    {
      "epoch": 0.7829752515637748,
      "grad_norm": 0.37890625,
      "learning_rate": 2.732843479736463e-06,
      "loss": 0.7012,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5758
    },
    {
      "epoch": 0.7831112319825945,
      "grad_norm": 0.67578125,
      "learning_rate": 2.7295828804125e-06,
      "loss": 0.8392,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5759
    },
    {
      "epoch": 0.7832472124014142,
      "grad_norm": 0.423828125,
      "learning_rate": 2.7263239199294477e-06,
      "loss": 0.7299,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5760
    },
    {
      "epoch": 0.7833831928202338,
      "grad_norm": 0.3828125,
      "learning_rate": 2.7230665990219084e-06,
      "loss": 0.7113,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5761
    },
    {
      "epoch": 0.7835191732390536,
      "grad_norm": 0.609375,
      "learning_rate": 2.7198109184241295e-06,
      "loss": 0.6267,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5762
    },
    {
      "epoch": 0.7836551536578733,
      "grad_norm": 0.4375,
      "learning_rate": 2.716556878869975e-06,
      "loss": 0.5907,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5763
    },
    {
      "epoch": 0.783791134076693,
      "grad_norm": 0.455078125,
      "learning_rate": 2.713304481092951e-06,
      "loss": 0.673,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5764
    },
    {
      "epoch": 0.7839271144955127,
      "grad_norm": 0.30859375,
      "learning_rate": 2.7100537258261805e-06,
      "loss": 0.6273,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5765
    },
    {
      "epoch": 0.7840630949143323,
      "grad_norm": 0.33984375,
      "learning_rate": 2.7068046138024318e-06,
      "loss": 0.6011,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5766
    },
    {
      "epoch": 0.784199075333152,
      "grad_norm": 0.77734375,
      "learning_rate": 2.7035571457540865e-06,
      "loss": 0.7391,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5767
    },
    {
      "epoch": 0.7843350557519717,
      "grad_norm": 0.400390625,
      "learning_rate": 2.7003113224131717e-06,
      "loss": 0.9141,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5768
    },
    {
      "epoch": 0.7844710361707914,
      "grad_norm": 0.478515625,
      "learning_rate": 2.6970671445113273e-06,
      "loss": 0.4375,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5769
    },
    {
      "epoch": 0.7846070165896111,
      "grad_norm": 0.40234375,
      "learning_rate": 2.6938246127798385e-06,
      "loss": 0.8101,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5770
    },
    {
      "epoch": 0.7847429970084308,
      "grad_norm": 0.32421875,
      "learning_rate": 2.690583727949604e-06,
      "loss": 0.6125,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5771
    },
    {
      "epoch": 0.7848789774272504,
      "grad_norm": 0.35546875,
      "learning_rate": 2.687344490751167e-06,
      "loss": 0.8044,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5772
    },
    {
      "epoch": 0.7850149578460701,
      "grad_norm": 0.388671875,
      "learning_rate": 2.6841069019146827e-06,
      "loss": 0.5695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5773
    },
    {
      "epoch": 0.7851509382648899,
      "grad_norm": 0.486328125,
      "learning_rate": 2.680870962169951e-06,
      "loss": 0.6038,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5774
    },
    {
      "epoch": 0.7852869186837096,
      "grad_norm": 0.41796875,
      "learning_rate": 2.6776366722463854e-06,
      "loss": 0.5231,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5775
    },
    {
      "epoch": 0.7854228991025293,
      "grad_norm": 0.3046875,
      "learning_rate": 2.6744040328730402e-06,
      "loss": 0.6335,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5776
    },
    {
      "epoch": 0.785558879521349,
      "grad_norm": 0.279296875,
      "learning_rate": 2.6711730447785876e-06,
      "loss": 0.5734,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5777
    },
    {
      "epoch": 0.7856948599401686,
      "grad_norm": 0.62109375,
      "learning_rate": 2.6679437086913352e-06,
      "loss": 0.7122,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5778
    },
    {
      "epoch": 0.7858308403589883,
      "grad_norm": 0.73046875,
      "learning_rate": 2.6647160253392114e-06,
      "loss": 0.7899,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5779
    },
    {
      "epoch": 0.785966820777808,
      "grad_norm": 0.345703125,
      "learning_rate": 2.6614899954497797e-06,
      "loss": 0.6258,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5780
    },
    {
      "epoch": 0.7861028011966277,
      "grad_norm": 0.36328125,
      "learning_rate": 2.658265619750221e-06,
      "loss": 0.5921,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5781
    },
    {
      "epoch": 0.7862387816154474,
      "grad_norm": 0.291015625,
      "learning_rate": 2.6550428989673548e-06,
      "loss": 0.5231,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5782
    },
    {
      "epoch": 0.786374762034267,
      "grad_norm": 0.458984375,
      "learning_rate": 2.651821833827617e-06,
      "loss": 0.7909,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5783
    },
    {
      "epoch": 0.7865107424530867,
      "grad_norm": 0.427734375,
      "learning_rate": 2.648602425057081e-06,
      "loss": 0.765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5784
    },
    {
      "epoch": 0.7866467228719064,
      "grad_norm": 0.28125,
      "learning_rate": 2.6453846733814324e-06,
      "loss": 0.4886,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5785
    },
    {
      "epoch": 0.7867827032907261,
      "grad_norm": 0.40234375,
      "learning_rate": 2.642168579526001e-06,
      "loss": 0.5892,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5786
    },
    {
      "epoch": 0.7869186837095459,
      "grad_norm": 0.408203125,
      "learning_rate": 2.638954144215725e-06,
      "loss": 0.7201,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5787
    },
    {
      "epoch": 0.7870546641283656,
      "grad_norm": 0.3671875,
      "learning_rate": 2.635741368175185e-06,
      "loss": 0.6816,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5788
    },
    {
      "epoch": 0.7871906445471852,
      "grad_norm": 0.255859375,
      "learning_rate": 2.6325302521285744e-06,
      "loss": 0.4399,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5789
    },
    {
      "epoch": 0.7873266249660049,
      "grad_norm": 0.62890625,
      "learning_rate": 2.6293207967997224e-06,
      "loss": 0.5002,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5790
    },
    {
      "epoch": 0.7874626053848246,
      "grad_norm": 0.341796875,
      "learning_rate": 2.626113002912075e-06,
      "loss": 0.6584,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5791
    },
    {
      "epoch": 0.7875985858036443,
      "grad_norm": 0.369140625,
      "learning_rate": 2.622906871188713e-06,
      "loss": 0.5721,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5792
    },
    {
      "epoch": 0.787734566222464,
      "grad_norm": 0.828125,
      "learning_rate": 2.619702402352332e-06,
      "loss": 0.7355,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5793
    },
    {
      "epoch": 0.7878705466412836,
      "grad_norm": 0.33203125,
      "learning_rate": 2.6164995971252647e-06,
      "loss": 0.6012,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5794
    },
    {
      "epoch": 0.7880065270601033,
      "grad_norm": 0.34375,
      "learning_rate": 2.6132984562294574e-06,
      "loss": 0.6584,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5795
    },
    {
      "epoch": 0.788142507478923,
      "grad_norm": 0.58203125,
      "learning_rate": 2.6100989803864896e-06,
      "loss": 0.5927,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5796
    },
    {
      "epoch": 0.7882784878977427,
      "grad_norm": 0.478515625,
      "learning_rate": 2.6069011703175596e-06,
      "loss": 0.5866,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5797
    },
    {
      "epoch": 0.7884144683165624,
      "grad_norm": 0.376953125,
      "learning_rate": 2.603705026743497e-06,
      "loss": 0.6228,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5798
    },
    {
      "epoch": 0.7885504487353822,
      "grad_norm": 0.296875,
      "learning_rate": 2.600510550384746e-06,
      "loss": 0.6549,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5799
    },
    {
      "epoch": 0.7886864291542018,
      "grad_norm": 0.33984375,
      "learning_rate": 2.597317741961387e-06,
      "loss": 0.6392,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5800
    },
    {
      "epoch": 0.7888224095730215,
      "grad_norm": 0.28515625,
      "learning_rate": 2.5941266021931112e-06,
      "loss": 0.5981,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5801
    },
    {
      "epoch": 0.7889583899918412,
      "grad_norm": 0.44921875,
      "learning_rate": 2.590937131799248e-06,
      "loss": 0.5882,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5802
    },
    {
      "epoch": 0.7890943704106609,
      "grad_norm": 0.3125,
      "learning_rate": 2.5877493314987357e-06,
      "loss": 0.5002,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5803
    },
    {
      "epoch": 0.7892303508294806,
      "grad_norm": 0.384765625,
      "learning_rate": 2.5845632020101495e-06,
      "loss": 0.744,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5804
    },
    {
      "epoch": 0.7893663312483002,
      "grad_norm": 0.3671875,
      "learning_rate": 2.5813787440516757e-06,
      "loss": 0.6147,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5805
    },
    {
      "epoch": 0.7895023116671199,
      "grad_norm": 0.4921875,
      "learning_rate": 2.5781959583411375e-06,
      "loss": 0.7775,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5806
    },
    {
      "epoch": 0.7896382920859396,
      "grad_norm": 0.482421875,
      "learning_rate": 2.5750148455959667e-06,
      "loss": 0.7931,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5807
    },
    {
      "epoch": 0.7897742725047593,
      "grad_norm": 0.61328125,
      "learning_rate": 2.571835406533232e-06,
      "loss": 0.5267,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5808
    },
    {
      "epoch": 0.789910252923579,
      "grad_norm": 0.5703125,
      "learning_rate": 2.56865764186961e-06,
      "loss": 0.6857,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5809
    },
    {
      "epoch": 0.7900462333423987,
      "grad_norm": 0.333984375,
      "learning_rate": 2.565481552321415e-06,
      "loss": 0.6605,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5810
    },
    {
      "epoch": 0.7901822137612183,
      "grad_norm": 0.36328125,
      "learning_rate": 2.5623071386045696e-06,
      "loss": 0.6657,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5811
    },
    {
      "epoch": 0.7903181941800381,
      "grad_norm": 0.34765625,
      "learning_rate": 2.559134401434632e-06,
      "loss": 0.658,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5812
    },
    {
      "epoch": 0.7904541745988578,
      "grad_norm": 0.39453125,
      "learning_rate": 2.5559633415267703e-06,
      "loss": 0.7061,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5813
    },
    {
      "epoch": 0.7905901550176775,
      "grad_norm": 0.51171875,
      "learning_rate": 2.552793959595786e-06,
      "loss": 0.7173,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5814
    },
    {
      "epoch": 0.7907261354364972,
      "grad_norm": 0.40625,
      "learning_rate": 2.54962625635609e-06,
      "loss": 0.7572,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5815
    },
    {
      "epoch": 0.7908621158553168,
      "grad_norm": 0.62109375,
      "learning_rate": 2.546460232521728e-06,
      "loss": 0.8021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5816
    },
    {
      "epoch": 0.7909980962741365,
      "grad_norm": 0.435546875,
      "learning_rate": 2.543295888806354e-06,
      "loss": 0.418,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5817
    },
    {
      "epoch": 0.7911340766929562,
      "grad_norm": 0.3359375,
      "learning_rate": 2.540133225923256e-06,
      "loss": 0.5835,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5818
    },
    {
      "epoch": 0.7912700571117759,
      "grad_norm": 0.314453125,
      "learning_rate": 2.536972244585331e-06,
      "loss": 0.5531,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5819
    },
    {
      "epoch": 0.7914060375305956,
      "grad_norm": 0.36328125,
      "learning_rate": 2.5338129455051053e-06,
      "loss": 0.6273,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5820
    },
    {
      "epoch": 0.7915420179494153,
      "grad_norm": 0.3828125,
      "learning_rate": 2.5306553293947265e-06,
      "loss": 0.7109,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5821
    },
    {
      "epoch": 0.7916779983682349,
      "grad_norm": 0.35546875,
      "learning_rate": 2.5274993969659554e-06,
      "loss": 0.7505,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5822
    },
    {
      "epoch": 0.7918139787870546,
      "grad_norm": 0.490234375,
      "learning_rate": 2.524345148930182e-06,
      "loss": 0.7378,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5823
    },
    {
      "epoch": 0.7919499592058743,
      "grad_norm": 0.3984375,
      "learning_rate": 2.5211925859984064e-06,
      "loss": 0.7106,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5824
    },
    {
      "epoch": 0.7920859396246941,
      "grad_norm": 0.33203125,
      "learning_rate": 2.5180417088812624e-06,
      "loss": 0.6551,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5825
    },
    {
      "epoch": 0.7922219200435138,
      "grad_norm": 0.56640625,
      "learning_rate": 2.514892518288988e-06,
      "loss": 0.613,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5826
    },
    {
      "epoch": 0.7923579004623335,
      "grad_norm": 0.48046875,
      "learning_rate": 2.511745014931458e-06,
      "loss": 0.8293,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5827
    },
    {
      "epoch": 0.7924938808811531,
      "grad_norm": 0.41796875,
      "learning_rate": 2.50859919951815e-06,
      "loss": 0.9686,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5828
    },
    {
      "epoch": 0.7926298612999728,
      "grad_norm": 0.328125,
      "learning_rate": 2.5054550727581763e-06,
      "loss": 0.6842,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5829
    },
    {
      "epoch": 0.7927658417187925,
      "grad_norm": 0.376953125,
      "learning_rate": 2.5023126353602545e-06,
      "loss": 0.4359,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5830
    },
    {
      "epoch": 0.7929018221376122,
      "grad_norm": 0.75390625,
      "learning_rate": 2.4991718880327342e-06,
      "loss": 0.9902,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5831
    },
    {
      "epoch": 0.7930378025564319,
      "grad_norm": 0.220703125,
      "learning_rate": 2.4960328314835746e-06,
      "loss": 0.4409,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5832
    },
    {
      "epoch": 0.7931737829752515,
      "grad_norm": 0.2578125,
      "learning_rate": 2.492895466420361e-06,
      "loss": 0.4007,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5833
    },
    {
      "epoch": 0.7933097633940712,
      "grad_norm": 0.3125,
      "learning_rate": 2.4897597935502894e-06,
      "loss": 0.6154,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5834
    },
    {
      "epoch": 0.7934457438128909,
      "grad_norm": 0.427734375,
      "learning_rate": 2.4866258135801837e-06,
      "loss": 0.6945,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5835
    },
    {
      "epoch": 0.7935817242317106,
      "grad_norm": 0.41796875,
      "learning_rate": 2.483493527216475e-06,
      "loss": 0.8447,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5836
    },
    {
      "epoch": 0.7937177046505304,
      "grad_norm": 0.30859375,
      "learning_rate": 2.480362935165227e-06,
      "loss": 0.6231,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5837
    },
    {
      "epoch": 0.7938536850693501,
      "grad_norm": 0.578125,
      "learning_rate": 2.4772340381321047e-06,
      "loss": 0.741,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5838
    },
    {
      "epoch": 0.7939896654881697,
      "grad_norm": 0.80078125,
      "learning_rate": 2.4741068368224084e-06,
      "loss": 0.7708,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5839
    },
    {
      "epoch": 0.7941256459069894,
      "grad_norm": 0.365234375,
      "learning_rate": 2.4709813319410403e-06,
      "loss": 0.554,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5840
    },
    {
      "epoch": 0.7942616263258091,
      "grad_norm": 0.4609375,
      "learning_rate": 2.467857524192533e-06,
      "loss": 0.8358,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5841
    },
    {
      "epoch": 0.7943976067446288,
      "grad_norm": 0.283203125,
      "learning_rate": 2.464735414281024e-06,
      "loss": 0.5163,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5842
    },
    {
      "epoch": 0.7945335871634485,
      "grad_norm": 0.5703125,
      "learning_rate": 2.4616150029102837e-06,
      "loss": 0.5275,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5843
    },
    {
      "epoch": 0.7946695675822681,
      "grad_norm": 0.373046875,
      "learning_rate": 2.458496290783684e-06,
      "loss": 0.6806,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5844
    },
    {
      "epoch": 0.7948055480010878,
      "grad_norm": 0.546875,
      "learning_rate": 2.455379278604226e-06,
      "loss": 0.8851,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5845
    },
    {
      "epoch": 0.7949415284199075,
      "grad_norm": 0.34765625,
      "learning_rate": 2.4522639670745165e-06,
      "loss": 0.6654,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5846
    },
    {
      "epoch": 0.7950775088387272,
      "grad_norm": 0.5546875,
      "learning_rate": 2.449150356896791e-06,
      "loss": 0.8477,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5847
    },
    {
      "epoch": 0.7952134892575469,
      "grad_norm": 0.384765625,
      "learning_rate": 2.4460384487728893e-06,
      "loss": 0.7126,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5848
    },
    {
      "epoch": 0.7953494696763665,
      "grad_norm": 0.419921875,
      "learning_rate": 2.44292824340428e-06,
      "loss": 0.6502,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5849
    },
    {
      "epoch": 0.7954854500951863,
      "grad_norm": 0.314453125,
      "learning_rate": 2.4398197414920345e-06,
      "loss": 0.5675,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5850
    },
    {
      "epoch": 0.795621430514006,
      "grad_norm": 0.38671875,
      "learning_rate": 2.4367129437368542e-06,
      "loss": 0.6852,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5851
    },
    {
      "epoch": 0.7957574109328257,
      "grad_norm": 0.2734375,
      "learning_rate": 2.4336078508390426e-06,
      "loss": 0.4997,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5852
    },
    {
      "epoch": 0.7958933913516454,
      "grad_norm": 0.734375,
      "learning_rate": 2.430504463498531e-06,
      "loss": 0.8504,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5853
    },
    {
      "epoch": 0.7960293717704651,
      "grad_norm": 0.6875,
      "learning_rate": 2.427402782414856e-06,
      "loss": 0.7287,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5854
    },
    {
      "epoch": 0.7961653521892847,
      "grad_norm": 0.396484375,
      "learning_rate": 2.4243028082871813e-06,
      "loss": 0.6497,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5855
    },
    {
      "epoch": 0.7963013326081044,
      "grad_norm": 0.578125,
      "learning_rate": 2.4212045418142706e-06,
      "loss": 0.7166,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5856
    },
    {
      "epoch": 0.7964373130269241,
      "grad_norm": 0.345703125,
      "learning_rate": 2.4181079836945175e-06,
      "loss": 0.6538,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5857
    },
    {
      "epoch": 0.7965732934457438,
      "grad_norm": 0.53515625,
      "learning_rate": 2.41501313462592e-06,
      "loss": 0.54,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5858
    },
    {
      "epoch": 0.7967092738645635,
      "grad_norm": 0.255859375,
      "learning_rate": 2.4119199953060977e-06,
      "loss": 0.394,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5859
    },
    {
      "epoch": 0.7968452542833832,
      "grad_norm": 0.5546875,
      "learning_rate": 2.4088285664322797e-06,
      "loss": 0.6484,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5860
    },
    {
      "epoch": 0.7969812347022028,
      "grad_norm": 0.392578125,
      "learning_rate": 2.4057388487013157e-06,
      "loss": 0.5418,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5861
    },
    {
      "epoch": 0.7971172151210226,
      "grad_norm": 0.494140625,
      "learning_rate": 2.4026508428096594e-06,
      "loss": 0.731,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5862
    },
    {
      "epoch": 0.7972531955398423,
      "grad_norm": 0.369140625,
      "learning_rate": 2.3995645494533925e-06,
      "loss": 0.7104,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5863
    },
    {
      "epoch": 0.797389175958662,
      "grad_norm": 0.578125,
      "learning_rate": 2.3964799693281972e-06,
      "loss": 0.8467,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5864
    },
    {
      "epoch": 0.7975251563774817,
      "grad_norm": 0.28125,
      "learning_rate": 2.39339710312938e-06,
      "loss": 0.4917,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5865
    },
    {
      "epoch": 0.7976611367963014,
      "grad_norm": 0.5390625,
      "learning_rate": 2.3903159515518527e-06,
      "loss": 0.6852,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5866
    },
    {
      "epoch": 0.797797117215121,
      "grad_norm": 0.5390625,
      "learning_rate": 2.387236515290149e-06,
      "loss": 0.5415,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5867
    },
    {
      "epoch": 0.7979330976339407,
      "grad_norm": 0.65625,
      "learning_rate": 2.384158795038407e-06,
      "loss": 0.705,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5868
    },
    {
      "epoch": 0.7980690780527604,
      "grad_norm": 0.396484375,
      "learning_rate": 2.3810827914903867e-06,
      "loss": 0.6644,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5869
    },
    {
      "epoch": 0.7982050584715801,
      "grad_norm": 0.345703125,
      "learning_rate": 2.3780085053394508e-06,
      "loss": 0.5859,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5870
    },
    {
      "epoch": 0.7983410388903998,
      "grad_norm": 0.32421875,
      "learning_rate": 2.3749359372785884e-06,
      "loss": 0.6206,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5871
    },
    {
      "epoch": 0.7984770193092194,
      "grad_norm": 0.38671875,
      "learning_rate": 2.371865088000388e-06,
      "loss": 0.7207,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5872
    },
    {
      "epoch": 0.7986129997280391,
      "grad_norm": 0.451171875,
      "learning_rate": 2.3687959581970623e-06,
      "loss": 0.6995,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5873
    },
    {
      "epoch": 0.7987489801468588,
      "grad_norm": 0.451171875,
      "learning_rate": 2.365728548560423e-06,
      "loss": 0.7366,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5874
    },
    {
      "epoch": 0.7988849605656786,
      "grad_norm": 0.384765625,
      "learning_rate": 2.36266285978191e-06,
      "loss": 0.7116,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5875
    },
    {
      "epoch": 0.7990209409844983,
      "grad_norm": 0.333984375,
      "learning_rate": 2.3595988925525594e-06,
      "loss": 0.6362,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5876
    },
    {
      "epoch": 0.799156921403318,
      "grad_norm": 0.259765625,
      "learning_rate": 2.3565366475630345e-06,
      "loss": 0.464,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5877
    },
    {
      "epoch": 0.7992929018221376,
      "grad_norm": 0.4140625,
      "learning_rate": 2.3534761255035955e-06,
      "loss": 0.6117,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5878
    },
    {
      "epoch": 0.7994288822409573,
      "grad_norm": 0.7265625,
      "learning_rate": 2.3504173270641274e-06,
      "loss": 0.6022,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5879
    },
    {
      "epoch": 0.799564862659777,
      "grad_norm": 0.99609375,
      "learning_rate": 2.3473602529341155e-06,
      "loss": 0.7179,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5880
    },
    {
      "epoch": 0.7997008430785967,
      "grad_norm": 0.2080078125,
      "learning_rate": 2.3443049038026643e-06,
      "loss": 0.3089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5881
    },
    {
      "epoch": 0.7998368234974164,
      "grad_norm": 0.361328125,
      "learning_rate": 2.3412512803584896e-06,
      "loss": 0.5708,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5882
    },
    {
      "epoch": 0.799972803916236,
      "grad_norm": 0.248046875,
      "learning_rate": 2.3381993832899096e-06,
      "loss": 0.4769,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5883
    },
    {
      "epoch": 0.8001087843350557,
      "grad_norm": 0.89453125,
      "learning_rate": 2.3351492132848665e-06,
      "loss": 0.7033,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5884
    },
    {
      "epoch": 0.8002447647538754,
      "grad_norm": 0.330078125,
      "learning_rate": 2.3321007710308985e-06,
      "loss": 0.5767,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5885
    },
    {
      "epoch": 0.8003807451726951,
      "grad_norm": 0.46875,
      "learning_rate": 2.3290540572151675e-06,
      "loss": 0.6701,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5886
    },
    {
      "epoch": 0.8005167255915149,
      "grad_norm": 0.498046875,
      "learning_rate": 2.3260090725244342e-06,
      "loss": 0.7882,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5887
    },
    {
      "epoch": 0.8006527060103346,
      "grad_norm": 0.41796875,
      "learning_rate": 2.322965817645083e-06,
      "loss": 0.6587,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5888
    },
    {
      "epoch": 0.8007886864291542,
      "grad_norm": 0.251953125,
      "learning_rate": 2.3199242932630926e-06,
      "loss": 0.5171,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5889
    },
    {
      "epoch": 0.8009246668479739,
      "grad_norm": 0.25390625,
      "learning_rate": 2.3168845000640673e-06,
      "loss": 0.5448,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5890
    },
    {
      "epoch": 0.8010606472667936,
      "grad_norm": 0.3125,
      "learning_rate": 2.3138464387332073e-06,
      "loss": 0.6501,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5891
    },
    {
      "epoch": 0.8011966276856133,
      "grad_norm": 0.64453125,
      "learning_rate": 2.310810109955334e-06,
      "loss": 0.6164,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5892
    },
    {
      "epoch": 0.801332608104433,
      "grad_norm": 0.396484375,
      "learning_rate": 2.307775514414867e-06,
      "loss": 0.66,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5893
    },
    {
      "epoch": 0.8014685885232526,
      "grad_norm": 0.255859375,
      "learning_rate": 2.3047426527958483e-06,
      "loss": 0.4328,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5894
    },
    {
      "epoch": 0.8016045689420723,
      "grad_norm": 0.3125,
      "learning_rate": 2.3017115257819157e-06,
      "loss": 0.5511,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5895
    },
    {
      "epoch": 0.801740549360892,
      "grad_norm": 0.271484375,
      "learning_rate": 2.298682134056328e-06,
      "loss": 0.5664,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5896
    },
    {
      "epoch": 0.8018765297797117,
      "grad_norm": 0.65234375,
      "learning_rate": 2.295654478301942e-06,
      "loss": 0.7529,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5897
    },
    {
      "epoch": 0.8020125101985314,
      "grad_norm": 0.357421875,
      "learning_rate": 2.292628559201232e-06,
      "loss": 0.7025,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5898
    },
    {
      "epoch": 0.802148490617351,
      "grad_norm": 0.2431640625,
      "learning_rate": 2.289604377436273e-06,
      "loss": 0.4697,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5899
    },
    {
      "epoch": 0.8022844710361708,
      "grad_norm": 0.36328125,
      "learning_rate": 2.2865819336887596e-06,
      "loss": 0.6056,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5900
    },
    {
      "epoch": 0.8024204514549905,
      "grad_norm": 0.3828125,
      "learning_rate": 2.2835612286399787e-06,
      "loss": 0.7371,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5901
    },
    {
      "epoch": 0.8025564318738102,
      "grad_norm": 0.279296875,
      "learning_rate": 2.2805422629708427e-06,
      "loss": 0.6336,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5902
    },
    {
      "epoch": 0.8026924122926299,
      "grad_norm": 0.341796875,
      "learning_rate": 2.2775250373618553e-06,
      "loss": 0.4852,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5903
    },
    {
      "epoch": 0.8028283927114496,
      "grad_norm": 0.408203125,
      "learning_rate": 2.2745095524931438e-06,
      "loss": 0.7183,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5904
    },
    {
      "epoch": 0.8029643731302692,
      "grad_norm": 0.55078125,
      "learning_rate": 2.2714958090444283e-06,
      "loss": 0.5269,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5905
    },
    {
      "epoch": 0.8031003535490889,
      "grad_norm": 1.0,
      "learning_rate": 2.26848380769505e-06,
      "loss": 0.6221,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5906
    },
    {
      "epoch": 0.8032363339679086,
      "grad_norm": 0.318359375,
      "learning_rate": 2.2654735491239445e-06,
      "loss": 0.57,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5907
    },
    {
      "epoch": 0.8033723143867283,
      "grad_norm": 0.3515625,
      "learning_rate": 2.2624650340096676e-06,
      "loss": 0.5439,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5908
    },
    {
      "epoch": 0.803508294805548,
      "grad_norm": 0.3984375,
      "learning_rate": 2.2594582630303674e-06,
      "loss": 0.6237,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5909
    },
    {
      "epoch": 0.8036442752243677,
      "grad_norm": 0.3515625,
      "learning_rate": 2.256453236863815e-06,
      "loss": 0.709,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5910
    },
    {
      "epoch": 0.8037802556431873,
      "grad_norm": 0.39453125,
      "learning_rate": 2.2534499561873737e-06,
      "loss": 0.6475,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5911
    },
    {
      "epoch": 0.803916236062007,
      "grad_norm": 0.328125,
      "learning_rate": 2.2504484216780243e-06,
      "loss": 0.6567,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5912
    },
    {
      "epoch": 0.8040522164808268,
      "grad_norm": 0.62890625,
      "learning_rate": 2.2474486340123445e-06,
      "loss": 0.7521,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5913
    },
    {
      "epoch": 0.8041881968996465,
      "grad_norm": 0.369140625,
      "learning_rate": 2.2444505938665296e-06,
      "loss": 0.8517,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5914
    },
    {
      "epoch": 0.8043241773184662,
      "grad_norm": 0.359375,
      "learning_rate": 2.241454301916367e-06,
      "loss": 0.6787,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5915
    },
    {
      "epoch": 0.8044601577372859,
      "grad_norm": 0.275390625,
      "learning_rate": 2.2384597588372647e-06,
      "loss": 0.4331,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5916
    },
    {
      "epoch": 0.8045961381561055,
      "grad_norm": 0.39453125,
      "learning_rate": 2.235466965304223e-06,
      "loss": 0.7809,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5917
    },
    {
      "epoch": 0.8047321185749252,
      "grad_norm": 0.90234375,
      "learning_rate": 2.232475921991859e-06,
      "loss": 0.598,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5918
    },
    {
      "epoch": 0.8048680989937449,
      "grad_norm": 0.384765625,
      "learning_rate": 2.2294866295743867e-06,
      "loss": 0.6779,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5919
    },
    {
      "epoch": 0.8050040794125646,
      "grad_norm": 0.298828125,
      "learning_rate": 2.2264990887256343e-06,
      "loss": 0.6219,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5920
    },
    {
      "epoch": 0.8051400598313843,
      "grad_norm": 0.6015625,
      "learning_rate": 2.2235133001190233e-06,
      "loss": 0.6605,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5921
    },
    {
      "epoch": 0.8052760402502039,
      "grad_norm": 0.341796875,
      "learning_rate": 2.2205292644275933e-06,
      "loss": 0.5368,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5922
    },
    {
      "epoch": 0.8054120206690236,
      "grad_norm": 0.56640625,
      "learning_rate": 2.217546982323977e-06,
      "loss": 0.8218,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5923
    },
    {
      "epoch": 0.8055480010878433,
      "grad_norm": 0.283203125,
      "learning_rate": 2.2145664544804223e-06,
      "loss": 0.5099,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5924
    },
    {
      "epoch": 0.8056839815066631,
      "grad_norm": 0.4609375,
      "learning_rate": 2.211587681568771e-06,
      "loss": 0.658,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5925
    },
    {
      "epoch": 0.8058199619254828,
      "grad_norm": 0.75390625,
      "learning_rate": 2.208610664260481e-06,
      "loss": 0.7166,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5926
    },
    {
      "epoch": 0.8059559423443025,
      "grad_norm": 0.345703125,
      "learning_rate": 2.2056354032266024e-06,
      "loss": 0.6313,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5927
    },
    {
      "epoch": 0.8060919227631221,
      "grad_norm": 0.265625,
      "learning_rate": 2.2026618991378026e-06,
      "loss": 0.4274,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5928
    },
    {
      "epoch": 0.8062279031819418,
      "grad_norm": 0.392578125,
      "learning_rate": 2.1996901526643366e-06,
      "loss": 0.7489,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5929
    },
    {
      "epoch": 0.8063638836007615,
      "grad_norm": 0.5859375,
      "learning_rate": 2.1967201644760816e-06,
      "loss": 0.7243,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5930
    },
    {
      "epoch": 0.8064998640195812,
      "grad_norm": 1.2265625,
      "learning_rate": 2.193751935242502e-06,
      "loss": 0.8335,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5931
    },
    {
      "epoch": 0.8066358444384009,
      "grad_norm": 0.4296875,
      "learning_rate": 2.190785465632679e-06,
      "loss": 0.755,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5932
    },
    {
      "epoch": 0.8067718248572205,
      "grad_norm": 0.349609375,
      "learning_rate": 2.1878207563152855e-06,
      "loss": 0.6014,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5933
    },
    {
      "epoch": 0.8069078052760402,
      "grad_norm": 0.359375,
      "learning_rate": 2.184857807958608e-06,
      "loss": 0.6994,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5934
    },
    {
      "epoch": 0.8070437856948599,
      "grad_norm": 0.369140625,
      "learning_rate": 2.181896621230527e-06,
      "loss": 0.6413,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5935
    },
    {
      "epoch": 0.8071797661136796,
      "grad_norm": 0.48828125,
      "learning_rate": 2.178937196798534e-06,
      "loss": 0.7024,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5936
    },
    {
      "epoch": 0.8073157465324993,
      "grad_norm": 0.373046875,
      "learning_rate": 2.1759795353297165e-06,
      "loss": 0.6174,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5937
    },
    {
      "epoch": 0.8074517269513191,
      "grad_norm": 0.828125,
      "learning_rate": 2.1730236374907708e-06,
      "loss": 0.9087,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5938
    },
    {
      "epoch": 0.8075877073701387,
      "grad_norm": 0.419921875,
      "learning_rate": 2.170069503947988e-06,
      "loss": 0.4826,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5939
    },
    {
      "epoch": 0.8077236877889584,
      "grad_norm": 0.361328125,
      "learning_rate": 2.167117135367267e-06,
      "loss": 0.742,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5940
    },
    {
      "epoch": 0.8078596682077781,
      "grad_norm": 0.26953125,
      "learning_rate": 2.1641665324141137e-06,
      "loss": 0.5509,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5941
    },
    {
      "epoch": 0.8079956486265978,
      "grad_norm": 0.46875,
      "learning_rate": 2.1612176957536222e-06,
      "loss": 0.6292,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5942
    },
    {
      "epoch": 0.8081316290454175,
      "grad_norm": 0.345703125,
      "learning_rate": 2.1582706260505015e-06,
      "loss": 0.6209,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5943
    },
    {
      "epoch": 0.8082676094642371,
      "grad_norm": 0.314453125,
      "learning_rate": 2.155325323969054e-06,
      "loss": 0.4469,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5944
    },
    {
      "epoch": 0.8084035898830568,
      "grad_norm": 0.34375,
      "learning_rate": 2.1523817901731904e-06,
      "loss": 0.7903,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5945
    },
    {
      "epoch": 0.8085395703018765,
      "grad_norm": 0.9140625,
      "learning_rate": 2.1494400253264134e-06,
      "loss": 0.8229,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5946
    },
    {
      "epoch": 0.8086755507206962,
      "grad_norm": 0.41796875,
      "learning_rate": 2.146500030091839e-06,
      "loss": 0.6043,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5947
    },
    {
      "epoch": 0.8088115311395159,
      "grad_norm": 0.25390625,
      "learning_rate": 2.143561805132174e-06,
      "loss": 0.4823,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5948
    },
    {
      "epoch": 0.8089475115583356,
      "grad_norm": 0.361328125,
      "learning_rate": 2.140625351109733e-06,
      "loss": 0.6935,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5949
    },
    {
      "epoch": 0.8090834919771553,
      "grad_norm": 0.40234375,
      "learning_rate": 2.1376906686864252e-06,
      "loss": 0.6837,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5950
    },
    {
      "epoch": 0.809219472395975,
      "grad_norm": 0.53125,
      "learning_rate": 2.1347577585237676e-06,
      "loss": 0.7075,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5951
    },
    {
      "epoch": 0.8093554528147947,
      "grad_norm": 0.365234375,
      "learning_rate": 2.131826621282871e-06,
      "loss": 0.536,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5952
    },
    {
      "epoch": 0.8094914332336144,
      "grad_norm": 0.357421875,
      "learning_rate": 2.1288972576244528e-06,
      "loss": 0.5506,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5953
    },
    {
      "epoch": 0.8096274136524341,
      "grad_norm": 0.55078125,
      "learning_rate": 2.1259696682088226e-06,
      "loss": 0.6755,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5954
    },
    {
      "epoch": 0.8097633940712538,
      "grad_norm": 0.5546875,
      "learning_rate": 2.1230438536959007e-06,
      "loss": 0.7616,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5955
    },
    {
      "epoch": 0.8098993744900734,
      "grad_norm": 0.66015625,
      "learning_rate": 2.120119814745196e-06,
      "loss": 0.6602,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5956
    },
    {
      "epoch": 0.8100353549088931,
      "grad_norm": 0.6015625,
      "learning_rate": 2.117197552015827e-06,
      "loss": 0.7995,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5957
    },
    {
      "epoch": 0.8101713353277128,
      "grad_norm": 0.53125,
      "learning_rate": 2.1142770661665037e-06,
      "loss": 0.665,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5958
    },
    {
      "epoch": 0.8103073157465325,
      "grad_norm": 0.427734375,
      "learning_rate": 2.111358357855544e-06,
      "loss": 0.797,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5959
    },
    {
      "epoch": 0.8104432961653522,
      "grad_norm": 0.671875,
      "learning_rate": 2.108441427740854e-06,
      "loss": 0.842,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5960
    },
    {
      "epoch": 0.8105792765841718,
      "grad_norm": 0.63671875,
      "learning_rate": 2.105526276479952e-06,
      "loss": 0.8405,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5961
    },
    {
      "epoch": 0.8107152570029915,
      "grad_norm": 0.337890625,
      "learning_rate": 2.1026129047299436e-06,
      "loss": 0.4661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5962
    },
    {
      "epoch": 0.8108512374218113,
      "grad_norm": 0.380859375,
      "learning_rate": 2.099701313147543e-06,
      "loss": 0.7139,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5963
    },
    {
      "epoch": 0.810987217840631,
      "grad_norm": 0.400390625,
      "learning_rate": 2.0967915023890527e-06,
      "loss": 0.6792,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5964
    },
    {
      "epoch": 0.8111231982594507,
      "grad_norm": 0.64453125,
      "learning_rate": 2.0938834731103862e-06,
      "loss": 0.7899,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5965
    },
    {
      "epoch": 0.8112591786782704,
      "grad_norm": 0.74609375,
      "learning_rate": 2.0909772259670434e-06,
      "loss": 0.4966,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5966
    },
    {
      "epoch": 0.81139515909709,
      "grad_norm": 0.5703125,
      "learning_rate": 2.088072761614133e-06,
      "loss": 0.7435,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5967
    },
    {
      "epoch": 0.8115311395159097,
      "grad_norm": 0.412109375,
      "learning_rate": 2.0851700807063504e-06,
      "loss": 0.7158,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5968
    },
    {
      "epoch": 0.8116671199347294,
      "grad_norm": 0.44140625,
      "learning_rate": 2.0822691838980015e-06,
      "loss": 0.6654,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5969
    },
    {
      "epoch": 0.8118031003535491,
      "grad_norm": 0.271484375,
      "learning_rate": 2.07937007184298e-06,
      "loss": 0.6343,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5970
    },
    {
      "epoch": 0.8119390807723688,
      "grad_norm": 0.431640625,
      "learning_rate": 2.076472745194784e-06,
      "loss": 0.7241,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5971
    },
    {
      "epoch": 0.8120750611911884,
      "grad_norm": 0.294921875,
      "learning_rate": 2.073577204606503e-06,
      "loss": 0.6642,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5972
    },
    {
      "epoch": 0.8122110416100081,
      "grad_norm": 0.578125,
      "learning_rate": 2.0706834507308317e-06,
      "loss": 0.8485,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5973
    },
    {
      "epoch": 0.8123470220288278,
      "grad_norm": 0.62109375,
      "learning_rate": 2.067791484220052e-06,
      "loss": 0.585,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5974
    },
    {
      "epoch": 0.8124830024476475,
      "grad_norm": 0.357421875,
      "learning_rate": 2.064901305726055e-06,
      "loss": 0.7581,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5975
    },
    {
      "epoch": 0.8126189828664673,
      "grad_norm": 0.33203125,
      "learning_rate": 2.0620129159003154e-06,
      "loss": 0.6169,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5976
    },
    {
      "epoch": 0.812754963285287,
      "grad_norm": 0.251953125,
      "learning_rate": 2.059126315393918e-06,
      "loss": 0.5176,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5977
    },
    {
      "epoch": 0.8128909437041066,
      "grad_norm": 0.50390625,
      "learning_rate": 2.056241504857531e-06,
      "loss": 0.7471,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5978
    },
    {
      "epoch": 0.8130269241229263,
      "grad_norm": 0.671875,
      "learning_rate": 2.0533584849414346e-06,
      "loss": 0.5571,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5979
    },
    {
      "epoch": 0.813162904541746,
      "grad_norm": 0.333984375,
      "learning_rate": 2.0504772562954876e-06,
      "loss": 0.5448,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5980
    },
    {
      "epoch": 0.8132988849605657,
      "grad_norm": 0.421875,
      "learning_rate": 2.0475978195691624e-06,
      "loss": 0.6589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5981
    },
    {
      "epoch": 0.8134348653793854,
      "grad_norm": 0.486328125,
      "learning_rate": 2.044720175411512e-06,
      "loss": 0.6509,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5982
    },
    {
      "epoch": 0.813570845798205,
      "grad_norm": 0.375,
      "learning_rate": 2.0418443244711993e-06,
      "loss": 0.728,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5983
    },
    {
      "epoch": 0.8137068262170247,
      "grad_norm": 0.470703125,
      "learning_rate": 2.0389702673964696e-06,
      "loss": 0.8385,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5984
    },
    {
      "epoch": 0.8138428066358444,
      "grad_norm": 0.3515625,
      "learning_rate": 2.0360980048351764e-06,
      "loss": 0.6211,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5985
    },
    {
      "epoch": 0.8139787870546641,
      "grad_norm": 0.314453125,
      "learning_rate": 2.0332275374347576e-06,
      "loss": 0.6514,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5986
    },
    {
      "epoch": 0.8141147674734838,
      "grad_norm": 0.4140625,
      "learning_rate": 2.0303588658422568e-06,
      "loss": 0.6746,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5987
    },
    {
      "epoch": 0.8142507478923036,
      "grad_norm": 0.306640625,
      "learning_rate": 2.0274919907043033e-06,
      "loss": 0.5485,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5988
    },
    {
      "epoch": 0.8143867283111232,
      "grad_norm": 0.326171875,
      "learning_rate": 2.02462691266713e-06,
      "loss": 0.6549,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5989
    },
    {
      "epoch": 0.8145227087299429,
      "grad_norm": 0.3203125,
      "learning_rate": 2.0217636323765553e-06,
      "loss": 0.5671,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5990
    },
    {
      "epoch": 0.8146586891487626,
      "grad_norm": 0.326171875,
      "learning_rate": 2.0189021504780025e-06,
      "loss": 0.4639,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5991
    },
    {
      "epoch": 0.8147946695675823,
      "grad_norm": 0.419921875,
      "learning_rate": 2.01604246761648e-06,
      "loss": 0.651,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5992
    },
    {
      "epoch": 0.814930649986402,
      "grad_norm": 0.349609375,
      "learning_rate": 2.0131845844366015e-06,
      "loss": 0.6084,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5993
    },
    {
      "epoch": 0.8150666304052216,
      "grad_norm": 0.4765625,
      "learning_rate": 2.0103285015825613e-06,
      "loss": 0.5679,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5994
    },
    {
      "epoch": 0.8152026108240413,
      "grad_norm": 0.357421875,
      "learning_rate": 2.0074742196981623e-06,
      "loss": 0.6599,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5995
    },
    {
      "epoch": 0.815338591242861,
      "grad_norm": 0.2392578125,
      "learning_rate": 2.0046217394267886e-06,
      "loss": 0.471,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5996
    },
    {
      "epoch": 0.8154745716616807,
      "grad_norm": 0.345703125,
      "learning_rate": 2.0017710614114306e-06,
      "loss": 0.5257,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5997
    },
    {
      "epoch": 0.8156105520805004,
      "grad_norm": 0.439453125,
      "learning_rate": 1.9989221862946585e-06,
      "loss": 0.6528,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5998
    },
    {
      "epoch": 0.81574653249932,
      "grad_norm": 0.69921875,
      "learning_rate": 1.9960751147186507e-06,
      "loss": 0.6434,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 5999
    },
    {
      "epoch": 0.8158825129181397,
      "grad_norm": 0.31640625,
      "learning_rate": 1.9932298473251656e-06,
      "loss": 0.6647,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6000
    },
    {
      "epoch": 0.8160184933369595,
      "grad_norm": 0.46875,
      "learning_rate": 1.990386384755565e-06,
      "loss": 0.8304,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6001
    },
    {
      "epoch": 0.8161544737557792,
      "grad_norm": 0.380859375,
      "learning_rate": 1.987544727650801e-06,
      "loss": 0.5522,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6002
    },
    {
      "epoch": 0.8162904541745989,
      "grad_norm": 0.27734375,
      "learning_rate": 1.9847048766514142e-06,
      "loss": 0.4897,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6003
    },
    {
      "epoch": 0.8164264345934186,
      "grad_norm": 0.3828125,
      "learning_rate": 1.9818668323975464e-06,
      "loss": 0.6361,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6004
    },
    {
      "epoch": 0.8165624150122383,
      "grad_norm": 0.384765625,
      "learning_rate": 1.9790305955289223e-06,
      "loss": 0.7635,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6005
    },
    {
      "epoch": 0.8166983954310579,
      "grad_norm": 0.474609375,
      "learning_rate": 1.9761961666848707e-06,
      "loss": 0.7002,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6006
    },
    {
      "epoch": 0.8168343758498776,
      "grad_norm": 0.3984375,
      "learning_rate": 1.973363546504299e-06,
      "loss": 0.799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6007
    },
    {
      "epoch": 0.8169703562686973,
      "grad_norm": 0.640625,
      "learning_rate": 1.970532735625721e-06,
      "loss": 0.7585,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6008
    },
    {
      "epoch": 0.817106336687517,
      "grad_norm": 0.48828125,
      "learning_rate": 1.967703734687231e-06,
      "loss": 0.7404,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6009
    },
    {
      "epoch": 0.8172423171063367,
      "grad_norm": 0.361328125,
      "learning_rate": 1.964876544326525e-06,
      "loss": 0.688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6010
    },
    {
      "epoch": 0.8173782975251563,
      "grad_norm": 0.423828125,
      "learning_rate": 1.9620511651808814e-06,
      "loss": 0.5493,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6011
    },
    {
      "epoch": 0.817514277943976,
      "grad_norm": 0.498046875,
      "learning_rate": 1.95922759788718e-06,
      "loss": 0.6248,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6012
    },
    {
      "epoch": 0.8176502583627958,
      "grad_norm": 0.328125,
      "learning_rate": 1.9564058430818832e-06,
      "loss": 0.5425,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6013
    },
    {
      "epoch": 0.8177862387816155,
      "grad_norm": 0.26953125,
      "learning_rate": 1.9535859014010525e-06,
      "loss": 0.4997,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6014
    },
    {
      "epoch": 0.8179222192004352,
      "grad_norm": 0.349609375,
      "learning_rate": 1.9507677734803345e-06,
      "loss": 0.626,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6015
    },
    {
      "epoch": 0.8180581996192549,
      "grad_norm": 0.275390625,
      "learning_rate": 1.9479514599549733e-06,
      "loss": 0.6587,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6016
    },
    {
      "epoch": 0.8181941800380745,
      "grad_norm": 0.34765625,
      "learning_rate": 1.945136961459795e-06,
      "loss": 0.5815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6017
    },
    {
      "epoch": 0.8183301604568942,
      "grad_norm": 0.412109375,
      "learning_rate": 1.9423242786292275e-06,
      "loss": 0.7941,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6018
    },
    {
      "epoch": 0.8184661408757139,
      "grad_norm": 0.2314453125,
      "learning_rate": 1.9395134120972794e-06,
      "loss": 0.3796,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6019
    },
    {
      "epoch": 0.8186021212945336,
      "grad_norm": 0.57421875,
      "learning_rate": 1.9367043624975577e-06,
      "loss": 0.8599,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6020
    },
    {
      "epoch": 0.8187381017133533,
      "grad_norm": 0.546875,
      "learning_rate": 1.9338971304632535e-06,
      "loss": 0.6247,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6021
    },
    {
      "epoch": 0.8188740821321729,
      "grad_norm": 0.640625,
      "learning_rate": 1.931091716627156e-06,
      "loss": 0.4657,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6022
    },
    {
      "epoch": 0.8190100625509926,
      "grad_norm": 0.373046875,
      "learning_rate": 1.928288121621633e-06,
      "loss": 0.6331,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6023
    },
    {
      "epoch": 0.8191460429698123,
      "grad_norm": 0.5625,
      "learning_rate": 1.9254863460786553e-06,
      "loss": 0.5778,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6024
    },
    {
      "epoch": 0.819282023388632,
      "grad_norm": 1.4140625,
      "learning_rate": 1.9226863906297734e-06,
      "loss": 0.9604,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6025
    },
    {
      "epoch": 0.8194180038074518,
      "grad_norm": 0.84765625,
      "learning_rate": 1.919888255906135e-06,
      "loss": 0.8306,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6026
    },
    {
      "epoch": 0.8195539842262715,
      "grad_norm": 0.625,
      "learning_rate": 1.9170919425384695e-06,
      "loss": 0.7569,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6027
    },
    {
      "epoch": 0.8196899646450911,
      "grad_norm": 0.59765625,
      "learning_rate": 1.9142974511571045e-06,
      "loss": 0.6746,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6028
    },
    {
      "epoch": 0.8198259450639108,
      "grad_norm": 0.220703125,
      "learning_rate": 1.9115047823919476e-06,
      "loss": 0.4834,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6029
    },
    {
      "epoch": 0.8199619254827305,
      "grad_norm": 0.6796875,
      "learning_rate": 1.908713936872506e-06,
      "loss": 0.6911,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6030
    },
    {
      "epoch": 0.8200979059015502,
      "grad_norm": 0.3828125,
      "learning_rate": 1.9059249152278657e-06,
      "loss": 0.592,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6031
    },
    {
      "epoch": 0.8202338863203699,
      "grad_norm": 0.57421875,
      "learning_rate": 1.9031377180867106e-06,
      "loss": 0.7534,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6032
    },
    {
      "epoch": 0.8203698667391895,
      "grad_norm": 0.81640625,
      "learning_rate": 1.9003523460773044e-06,
      "loss": 0.8426,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6033
    },
    {
      "epoch": 0.8205058471580092,
      "grad_norm": 1.21875,
      "learning_rate": 1.8975687998275093e-06,
      "loss": 0.7344,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6034
    },
    {
      "epoch": 0.8206418275768289,
      "grad_norm": 0.30859375,
      "learning_rate": 1.8947870799647638e-06,
      "loss": 0.6517,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6035
    },
    {
      "epoch": 0.8207778079956486,
      "grad_norm": 0.447265625,
      "learning_rate": 1.8920071871161093e-06,
      "loss": 0.9225,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6036
    },
    {
      "epoch": 0.8209137884144683,
      "grad_norm": 0.353515625,
      "learning_rate": 1.8892291219081605e-06,
      "loss": 0.6087,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6037
    },
    {
      "epoch": 0.821049768833288,
      "grad_norm": 0.3359375,
      "learning_rate": 1.8864528849671338e-06,
      "loss": 0.5166,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6038
    },
    {
      "epoch": 0.8211857492521077,
      "grad_norm": 0.345703125,
      "learning_rate": 1.883678476918821e-06,
      "loss": 0.5173,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6039
    },
    {
      "epoch": 0.8213217296709274,
      "grad_norm": 0.248046875,
      "learning_rate": 1.880905898388612e-06,
      "loss": 0.4884,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6040
    },
    {
      "epoch": 0.8214577100897471,
      "grad_norm": 0.435546875,
      "learning_rate": 1.8781351500014767e-06,
      "loss": 0.8859,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6041
    },
    {
      "epoch": 0.8215936905085668,
      "grad_norm": 0.33984375,
      "learning_rate": 1.8753662323819788e-06,
      "loss": 0.5829,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6042
    },
    {
      "epoch": 0.8217296709273865,
      "grad_norm": 0.314453125,
      "learning_rate": 1.8725991461542624e-06,
      "loss": 0.474,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6043
    },
    {
      "epoch": 0.8218656513462061,
      "grad_norm": 0.3671875,
      "learning_rate": 1.8698338919420667e-06,
      "loss": 0.5422,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6044
    },
    {
      "epoch": 0.8220016317650258,
      "grad_norm": 0.28515625,
      "learning_rate": 1.8670704703687104e-06,
      "loss": 0.631,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6045
    },
    {
      "epoch": 0.8221376121838455,
      "grad_norm": 0.36328125,
      "learning_rate": 1.8643088820571054e-06,
      "loss": 0.5537,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6046
    },
    {
      "epoch": 0.8222735926026652,
      "grad_norm": 0.361328125,
      "learning_rate": 1.8615491276297437e-06,
      "loss": 0.7686,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6047
    },
    {
      "epoch": 0.8224095730214849,
      "grad_norm": 1.078125,
      "learning_rate": 1.8587912077087134e-06,
      "loss": 0.8219,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6048
    },
    {
      "epoch": 0.8225455534403046,
      "grad_norm": 0.66015625,
      "learning_rate": 1.856035122915677e-06,
      "loss": 0.6667,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6049
    },
    {
      "epoch": 0.8226815338591242,
      "grad_norm": 0.58203125,
      "learning_rate": 1.8532808738718943e-06,
      "loss": 0.7224,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6050
    },
    {
      "epoch": 0.822817514277944,
      "grad_norm": 0.283203125,
      "learning_rate": 1.8505284611982033e-06,
      "loss": 0.6017,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6051
    },
    {
      "epoch": 0.8229534946967637,
      "grad_norm": 0.5234375,
      "learning_rate": 1.8477778855150353e-06,
      "loss": 0.6848,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6052
    },
    {
      "epoch": 0.8230894751155834,
      "grad_norm": 0.328125,
      "learning_rate": 1.8450291474423999e-06,
      "loss": 0.6355,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6053
    },
    {
      "epoch": 0.8232254555344031,
      "grad_norm": 0.37109375,
      "learning_rate": 1.8422822475999003e-06,
      "loss": 0.5618,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6054
    },
    {
      "epoch": 0.8233614359532228,
      "grad_norm": 0.76953125,
      "learning_rate": 1.8395371866067159e-06,
      "loss": 0.8423,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6055
    },
    {
      "epoch": 0.8234974163720424,
      "grad_norm": 0.1904296875,
      "learning_rate": 1.8367939650816224e-06,
      "loss": 0.3128,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6056
    },
    {
      "epoch": 0.8236333967908621,
      "grad_norm": 0.255859375,
      "learning_rate": 1.8340525836429713e-06,
      "loss": 0.5293,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6057
    },
    {
      "epoch": 0.8237693772096818,
      "grad_norm": 0.4296875,
      "learning_rate": 1.8313130429087078e-06,
      "loss": 0.6473,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6058
    },
    {
      "epoch": 0.8239053576285015,
      "grad_norm": 0.3515625,
      "learning_rate": 1.8285753434963539e-06,
      "loss": 0.5841,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6059
    },
    {
      "epoch": 0.8240413380473212,
      "grad_norm": 0.28515625,
      "learning_rate": 1.82583948602302e-06,
      "loss": 0.6571,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6060
    },
    {
      "epoch": 0.8241773184661408,
      "grad_norm": 0.609375,
      "learning_rate": 1.8231054711054075e-06,
      "loss": 0.6893,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6061
    },
    {
      "epoch": 0.8243132988849605,
      "grad_norm": 0.208984375,
      "learning_rate": 1.8203732993597912e-06,
      "loss": 0.4012,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6062
    },
    {
      "epoch": 0.8244492793037802,
      "grad_norm": 0.359375,
      "learning_rate": 1.8176429714020394e-06,
      "loss": 0.6729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6063
    },
    {
      "epoch": 0.8245852597226,
      "grad_norm": 0.53515625,
      "learning_rate": 1.8149144878475989e-06,
      "loss": 0.6268,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6064
    },
    {
      "epoch": 0.8247212401414197,
      "grad_norm": 0.2412109375,
      "learning_rate": 1.8121878493115053e-06,
      "loss": 0.4741,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6065
    },
    {
      "epoch": 0.8248572205602394,
      "grad_norm": 0.419921875,
      "learning_rate": 1.8094630564083737e-06,
      "loss": 0.5539,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6066
    },
    {
      "epoch": 0.824993200979059,
      "grad_norm": 0.421875,
      "learning_rate": 1.8067401097524095e-06,
      "loss": 0.5433,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6067
    },
    {
      "epoch": 0.8251291813978787,
      "grad_norm": 0.369140625,
      "learning_rate": 1.804019009957394e-06,
      "loss": 0.6001,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6068
    },
    {
      "epoch": 0.8252651618166984,
      "grad_norm": 0.59375,
      "learning_rate": 1.8012997576367008e-06,
      "loss": 0.8298,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6069
    },
    {
      "epoch": 0.8254011422355181,
      "grad_norm": 0.326171875,
      "learning_rate": 1.7985823534032775e-06,
      "loss": 0.6254,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6070
    },
    {
      "epoch": 0.8255371226543378,
      "grad_norm": 0.390625,
      "learning_rate": 1.7958667978696654e-06,
      "loss": 0.6611,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6071
    },
    {
      "epoch": 0.8256731030731574,
      "grad_norm": 0.76953125,
      "learning_rate": 1.793153091647979e-06,
      "loss": 0.7155,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6072
    },
    {
      "epoch": 0.8258090834919771,
      "grad_norm": 0.470703125,
      "learning_rate": 1.7904412353499256e-06,
      "loss": 0.5961,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6073
    },
    {
      "epoch": 0.8259450639107968,
      "grad_norm": 0.77734375,
      "learning_rate": 1.7877312295867854e-06,
      "loss": 0.5871,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6074
    },
    {
      "epoch": 0.8260810443296165,
      "grad_norm": 0.380859375,
      "learning_rate": 1.7850230749694331e-06,
      "loss": 0.7272,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6075
    },
    {
      "epoch": 0.8262170247484363,
      "grad_norm": 0.77734375,
      "learning_rate": 1.7823167721083145e-06,
      "loss": 0.8006,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6076
    },
    {
      "epoch": 0.826353005167256,
      "grad_norm": 0.3515625,
      "learning_rate": 1.7796123216134676e-06,
      "loss": 0.5994,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6077
    },
    {
      "epoch": 0.8264889855860756,
      "grad_norm": 0.486328125,
      "learning_rate": 1.7769097240945043e-06,
      "loss": 0.828,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6078
    },
    {
      "epoch": 0.8266249660048953,
      "grad_norm": 0.376953125,
      "learning_rate": 1.7742089801606278e-06,
      "loss": 0.7899,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6079
    },
    {
      "epoch": 0.826760946423715,
      "grad_norm": 0.349609375,
      "learning_rate": 1.771510090420613e-06,
      "loss": 0.7406,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6080
    },
    {
      "epoch": 0.8268969268425347,
      "grad_norm": 0.37109375,
      "learning_rate": 1.7688130554828297e-06,
      "loss": 0.6947,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6081
    },
    {
      "epoch": 0.8270329072613544,
      "grad_norm": 0.546875,
      "learning_rate": 1.7661178759552156e-06,
      "loss": 0.625,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6082
    },
    {
      "epoch": 0.827168887680174,
      "grad_norm": 0.255859375,
      "learning_rate": 1.763424552445303e-06,
      "loss": 0.5521,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6083
    },
    {
      "epoch": 0.8273048680989937,
      "grad_norm": 0.66015625,
      "learning_rate": 1.7607330855601945e-06,
      "loss": 0.7264,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6084
    },
    {
      "epoch": 0.8274408485178134,
      "grad_norm": 0.455078125,
      "learning_rate": 1.7580434759065857e-06,
      "loss": 0.4012,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6085
    },
    {
      "epoch": 0.8275768289366331,
      "grad_norm": 0.353515625,
      "learning_rate": 1.7553557240907414e-06,
      "loss": 0.7262,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6086
    },
    {
      "epoch": 0.8277128093554528,
      "grad_norm": 0.408203125,
      "learning_rate": 1.7526698307185186e-06,
      "loss": 0.8055,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6087
    },
    {
      "epoch": 0.8278487897742725,
      "grad_norm": 0.419921875,
      "learning_rate": 1.749985796395347e-06,
      "loss": 0.6789,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6088
    },
    {
      "epoch": 0.8279847701930922,
      "grad_norm": 0.5234375,
      "learning_rate": 1.747303621726244e-06,
      "loss": 0.5895,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6089
    },
    {
      "epoch": 0.8281207506119119,
      "grad_norm": 0.32421875,
      "learning_rate": 1.7446233073158003e-06,
      "loss": 0.4627,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6090
    },
    {
      "epoch": 0.8282567310307316,
      "grad_norm": 0.29296875,
      "learning_rate": 1.741944853768196e-06,
      "loss": 0.5332,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6091
    },
    {
      "epoch": 0.8283927114495513,
      "grad_norm": 0.3828125,
      "learning_rate": 1.7392682616871836e-06,
      "loss": 0.6322,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6092
    },
    {
      "epoch": 0.828528691868371,
      "grad_norm": 0.3203125,
      "learning_rate": 1.736593531676104e-06,
      "loss": 0.5063,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6093
    },
    {
      "epoch": 0.8286646722871907,
      "grad_norm": 0.328125,
      "learning_rate": 1.7339206643378692e-06,
      "loss": 0.5324,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6094
    },
    {
      "epoch": 0.8288006527060103,
      "grad_norm": 0.349609375,
      "learning_rate": 1.7312496602749795e-06,
      "loss": 0.5697,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6095
    },
    {
      "epoch": 0.82893663312483,
      "grad_norm": 0.427734375,
      "learning_rate": 1.7285805200895088e-06,
      "loss": 0.8109,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6096
    },
    {
      "epoch": 0.8290726135436497,
      "grad_norm": 0.3984375,
      "learning_rate": 1.7259132443831194e-06,
      "loss": 0.6464,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6097
    },
    {
      "epoch": 0.8292085939624694,
      "grad_norm": 0.279296875,
      "learning_rate": 1.7232478337570413e-06,
      "loss": 0.5197,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6098
    },
    {
      "epoch": 0.8293445743812891,
      "grad_norm": 0.6171875,
      "learning_rate": 1.720584288812096e-06,
      "loss": 0.62,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6099
    },
    {
      "epoch": 0.8294805548001087,
      "grad_norm": 0.3359375,
      "learning_rate": 1.7179226101486746e-06,
      "loss": 0.4813,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6100
    },
    {
      "epoch": 0.8296165352189284,
      "grad_norm": 0.28515625,
      "learning_rate": 1.7152627983667558e-06,
      "loss": 0.5682,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6101
    },
    {
      "epoch": 0.8297525156377482,
      "grad_norm": 0.51953125,
      "learning_rate": 1.7126048540658913e-06,
      "loss": 0.5181,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6102
    },
    {
      "epoch": 0.8298884960565679,
      "grad_norm": 0.40234375,
      "learning_rate": 1.7099487778452162e-06,
      "loss": 0.5786,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6103
    },
    {
      "epoch": 0.8300244764753876,
      "grad_norm": 0.4765625,
      "learning_rate": 1.7072945703034393e-06,
      "loss": 0.7336,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6104
    },
    {
      "epoch": 0.8301604568942073,
      "grad_norm": 0.2275390625,
      "learning_rate": 1.7046422320388556e-06,
      "loss": 0.4385,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6105
    },
    {
      "epoch": 0.8302964373130269,
      "grad_norm": 0.404296875,
      "learning_rate": 1.701991763649331e-06,
      "loss": 0.8257,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6106
    },
    {
      "epoch": 0.8304324177318466,
      "grad_norm": 0.55859375,
      "learning_rate": 1.6993431657323167e-06,
      "loss": 0.8289,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6107
    },
    {
      "epoch": 0.8305683981506663,
      "grad_norm": 0.484375,
      "learning_rate": 1.696696438884835e-06,
      "loss": 0.3883,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6108
    },
    {
      "epoch": 0.830704378569486,
      "grad_norm": 0.2734375,
      "learning_rate": 1.6940515837034954e-06,
      "loss": 0.5664,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6109
    },
    {
      "epoch": 0.8308403589883057,
      "grad_norm": 0.67578125,
      "learning_rate": 1.6914086007844754e-06,
      "loss": 0.6776,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6110
    },
    {
      "epoch": 0.8309763394071253,
      "grad_norm": 0.49609375,
      "learning_rate": 1.6887674907235407e-06,
      "loss": 0.7702,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6111
    },
    {
      "epoch": 0.831112319825945,
      "grad_norm": 0.1865234375,
      "learning_rate": 1.6861282541160251e-06,
      "loss": 0.3193,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6112
    },
    {
      "epoch": 0.8312483002447647,
      "grad_norm": 0.30859375,
      "learning_rate": 1.6834908915568494e-06,
      "loss": 0.5352,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6113
    },
    {
      "epoch": 0.8313842806635845,
      "grad_norm": 0.5390625,
      "learning_rate": 1.6808554036405024e-06,
      "loss": 0.7896,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6114
    },
    {
      "epoch": 0.8315202610824042,
      "grad_norm": 0.37109375,
      "learning_rate": 1.6782217909610609e-06,
      "loss": 0.6965,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6115
    },
    {
      "epoch": 0.8316562415012239,
      "grad_norm": 0.6171875,
      "learning_rate": 1.6755900541121673e-06,
      "loss": 0.6302,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6116
    },
    {
      "epoch": 0.8317922219200435,
      "grad_norm": 0.34765625,
      "learning_rate": 1.6729601936870532e-06,
      "loss": 0.6938,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6117
    },
    {
      "epoch": 0.8319282023388632,
      "grad_norm": 0.39453125,
      "learning_rate": 1.6703322102785168e-06,
      "loss": 0.8035,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6118
    },
    {
      "epoch": 0.8320641827576829,
      "grad_norm": 0.4140625,
      "learning_rate": 1.667706104478941e-06,
      "loss": 0.8099,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6119
    },
    {
      "epoch": 0.8322001631765026,
      "grad_norm": 1.1484375,
      "learning_rate": 1.665081876880278e-06,
      "loss": 0.6377,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6120
    },
    {
      "epoch": 0.8323361435953223,
      "grad_norm": 0.328125,
      "learning_rate": 1.662459528074064e-06,
      "loss": 0.5768,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6121
    },
    {
      "epoch": 0.832472124014142,
      "grad_norm": 0.6875,
      "learning_rate": 1.659839058651409e-06,
      "loss": 0.8504,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6122
    },
    {
      "epoch": 0.8326081044329616,
      "grad_norm": 0.419921875,
      "learning_rate": 1.657220469202997e-06,
      "loss": 0.63,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6123
    },
    {
      "epoch": 0.8327440848517813,
      "grad_norm": 0.49609375,
      "learning_rate": 1.6546037603190922e-06,
      "loss": 0.7985,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6124
    },
    {
      "epoch": 0.832880065270601,
      "grad_norm": 0.4140625,
      "learning_rate": 1.651988932589529e-06,
      "loss": 0.6035,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6125
    },
    {
      "epoch": 0.8330160456894207,
      "grad_norm": 0.298828125,
      "learning_rate": 1.6493759866037263e-06,
      "loss": 0.6226,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6126
    },
    {
      "epoch": 0.8331520261082405,
      "grad_norm": 0.361328125,
      "learning_rate": 1.6467649229506688e-06,
      "loss": 0.6287,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6127
    },
    {
      "epoch": 0.8332880065270601,
      "grad_norm": 0.53125,
      "learning_rate": 1.644155742218927e-06,
      "loss": 0.5755,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6128
    },
    {
      "epoch": 0.8334239869458798,
      "grad_norm": 0.361328125,
      "learning_rate": 1.641548444996639e-06,
      "loss": 0.6646,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6129
    },
    {
      "epoch": 0.8335599673646995,
      "grad_norm": 0.578125,
      "learning_rate": 1.6389430318715239e-06,
      "loss": 0.5446,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6130
    },
    {
      "epoch": 0.8336959477835192,
      "grad_norm": 0.765625,
      "learning_rate": 1.6363395034308704e-06,
      "loss": 0.9019,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6131
    },
    {
      "epoch": 0.8338319282023389,
      "grad_norm": 0.30078125,
      "learning_rate": 1.6337378602615505e-06,
      "loss": 0.6268,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6132
    },
    {
      "epoch": 0.8339679086211585,
      "grad_norm": 0.3515625,
      "learning_rate": 1.6311381029500007e-06,
      "loss": 0.7847,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6133
    },
    {
      "epoch": 0.8341038890399782,
      "grad_norm": 0.412109375,
      "learning_rate": 1.628540232082243e-06,
      "loss": 0.6143,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6134
    },
    {
      "epoch": 0.8342398694587979,
      "grad_norm": 0.439453125,
      "learning_rate": 1.6259442482438648e-06,
      "loss": 0.5942,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6135
    },
    {
      "epoch": 0.8343758498776176,
      "grad_norm": 0.330078125,
      "learning_rate": 1.6233501520200378e-06,
      "loss": 0.5578,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6136
    },
    {
      "epoch": 0.8345118302964373,
      "grad_norm": 0.287109375,
      "learning_rate": 1.6207579439954969e-06,
      "loss": 0.555,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6137
    },
    {
      "epoch": 0.834647810715257,
      "grad_norm": 0.3203125,
      "learning_rate": 1.618167624754563e-06,
      "loss": 0.6126,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6138
    },
    {
      "epoch": 0.8347837911340767,
      "grad_norm": 0.365234375,
      "learning_rate": 1.6155791948811217e-06,
      "loss": 0.637,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6139
    },
    {
      "epoch": 0.8349197715528964,
      "grad_norm": 0.71484375,
      "learning_rate": 1.6129926549586417e-06,
      "loss": 0.6037,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6140
    },
    {
      "epoch": 0.8350557519717161,
      "grad_norm": 0.36328125,
      "learning_rate": 1.6104080055701531e-06,
      "loss": 0.4766,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6141
    },
    {
      "epoch": 0.8351917323905358,
      "grad_norm": 0.388671875,
      "learning_rate": 1.6078252472982758e-06,
      "loss": 0.8029,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6142
    },
    {
      "epoch": 0.8353277128093555,
      "grad_norm": 0.74609375,
      "learning_rate": 1.6052443807251883e-06,
      "loss": 0.6532,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6143
    },
    {
      "epoch": 0.8354636932281752,
      "grad_norm": 0.44921875,
      "learning_rate": 1.6026654064326553e-06,
      "loss": 0.4847,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6144
    },
    {
      "epoch": 0.8355996736469948,
      "grad_norm": 0.390625,
      "learning_rate": 1.6000883250020026e-06,
      "loss": 0.5566,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6145
    },
    {
      "epoch": 0.8357356540658145,
      "grad_norm": 0.283203125,
      "learning_rate": 1.5975131370141428e-06,
      "loss": 0.4728,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6146
    },
    {
      "epoch": 0.8358716344846342,
      "grad_norm": 0.365234375,
      "learning_rate": 1.5949398430495478e-06,
      "loss": 0.658,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6147
    },
    {
      "epoch": 0.8360076149034539,
      "grad_norm": 0.55078125,
      "learning_rate": 1.592368443688277e-06,
      "loss": 0.763,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6148
    },
    {
      "epoch": 0.8361435953222736,
      "grad_norm": 0.8203125,
      "learning_rate": 1.5897989395099478e-06,
      "loss": 0.8394,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6149
    },
    {
      "epoch": 0.8362795757410932,
      "grad_norm": 0.76171875,
      "learning_rate": 1.5872313310937647e-06,
      "loss": 0.7567,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6150
    },
    {
      "epoch": 0.8364155561599129,
      "grad_norm": 0.44140625,
      "learning_rate": 1.5846656190184917e-06,
      "loss": 0.7301,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6151
    },
    {
      "epoch": 0.8365515365787327,
      "grad_norm": 0.404296875,
      "learning_rate": 1.5821018038624768e-06,
      "loss": 0.5689,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6152
    },
    {
      "epoch": 0.8366875169975524,
      "grad_norm": 1.03125,
      "learning_rate": 1.5795398862036315e-06,
      "loss": 0.6006,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6153
    },
    {
      "epoch": 0.8368234974163721,
      "grad_norm": 0.494140625,
      "learning_rate": 1.576979866619448e-06,
      "loss": 0.5496,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6154
    },
    {
      "epoch": 0.8369594778351918,
      "grad_norm": 0.2421875,
      "learning_rate": 1.5744217456869803e-06,
      "loss": 0.494,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6155
    },
    {
      "epoch": 0.8370954582540114,
      "grad_norm": 0.345703125,
      "learning_rate": 1.571865523982865e-06,
      "loss": 0.6626,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6156
    },
    {
      "epoch": 0.8372314386728311,
      "grad_norm": 0.58984375,
      "learning_rate": 1.5693112020833012e-06,
      "loss": 0.6183,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6157
    },
    {
      "epoch": 0.8373674190916508,
      "grad_norm": 0.296875,
      "learning_rate": 1.5667587805640682e-06,
      "loss": 0.6257,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6158
    },
    {
      "epoch": 0.8375033995104705,
      "grad_norm": 0.5390625,
      "learning_rate": 1.5642082600005104e-06,
      "loss": 0.7176,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6159
    },
    {
      "epoch": 0.8376393799292902,
      "grad_norm": 0.59375,
      "learning_rate": 1.56165964096755e-06,
      "loss": 0.6727,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6160
    },
    {
      "epoch": 0.8377753603481098,
      "grad_norm": 0.455078125,
      "learning_rate": 1.5591129240396717e-06,
      "loss": 0.473,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6161
    },
    {
      "epoch": 0.8379113407669295,
      "grad_norm": 0.330078125,
      "learning_rate": 1.5565681097909413e-06,
      "loss": 0.3494,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6162
    },
    {
      "epoch": 0.8380473211857492,
      "grad_norm": 0.337890625,
      "learning_rate": 1.5540251987949872e-06,
      "loss": 0.6593,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6163
    },
    {
      "epoch": 0.8381833016045689,
      "grad_norm": 0.470703125,
      "learning_rate": 1.5514841916250157e-06,
      "loss": 0.7886,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6164
    },
    {
      "epoch": 0.8383192820233887,
      "grad_norm": 0.5546875,
      "learning_rate": 1.5489450888537982e-06,
      "loss": 0.7267,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6165
    },
    {
      "epoch": 0.8384552624422084,
      "grad_norm": 0.46484375,
      "learning_rate": 1.546407891053684e-06,
      "loss": 0.6509,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6166
    },
    {
      "epoch": 0.838591242861028,
      "grad_norm": 0.64453125,
      "learning_rate": 1.5438725987965818e-06,
      "loss": 0.5929,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6167
    },
    {
      "epoch": 0.8387272232798477,
      "grad_norm": 0.484375,
      "learning_rate": 1.5413392126539839e-06,
      "loss": 0.5283,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6168
    },
    {
      "epoch": 0.8388632036986674,
      "grad_norm": 0.314453125,
      "learning_rate": 1.5388077331969409e-06,
      "loss": 0.6007,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6169
    },
    {
      "epoch": 0.8389991841174871,
      "grad_norm": 0.38671875,
      "learning_rate": 1.5362781609960853e-06,
      "loss": 0.6912,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6170
    },
    {
      "epoch": 0.8391351645363068,
      "grad_norm": 0.828125,
      "learning_rate": 1.5337504966216077e-06,
      "loss": 0.7435,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6171
    },
    {
      "epoch": 0.8392711449551264,
      "grad_norm": 0.5859375,
      "learning_rate": 1.531224740643279e-06,
      "loss": 0.4271,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6172
    },
    {
      "epoch": 0.8394071253739461,
      "grad_norm": 0.578125,
      "learning_rate": 1.5287008936304314e-06,
      "loss": 0.6475,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6173
    },
    {
      "epoch": 0.8395431057927658,
      "grad_norm": 0.470703125,
      "learning_rate": 1.5261789561519768e-06,
      "loss": 0.828,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6174
    },
    {
      "epoch": 0.8396790862115855,
      "grad_norm": 0.5625,
      "learning_rate": 1.5236589287763836e-06,
      "loss": 0.5889,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6175
    },
    {
      "epoch": 0.8398150666304052,
      "grad_norm": 0.451171875,
      "learning_rate": 1.521140812071703e-06,
      "loss": 0.7739,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6176
    },
    {
      "epoch": 0.839951047049225,
      "grad_norm": 1.0,
      "learning_rate": 1.5186246066055465e-06,
      "loss": 0.7912,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6177
    },
    {
      "epoch": 0.8400870274680446,
      "grad_norm": 0.306640625,
      "learning_rate": 1.516110312945095e-06,
      "loss": 0.6709,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6178
    },
    {
      "epoch": 0.8402230078868643,
      "grad_norm": 0.41015625,
      "learning_rate": 1.513597931657106e-06,
      "loss": 0.6545,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6179
    },
    {
      "epoch": 0.840358988305684,
      "grad_norm": 0.57421875,
      "learning_rate": 1.5110874633078943e-06,
      "loss": 0.6322,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6180
    },
    {
      "epoch": 0.8404949687245037,
      "grad_norm": 0.29296875,
      "learning_rate": 1.5085789084633596e-06,
      "loss": 0.5518,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6181
    },
    {
      "epoch": 0.8406309491433234,
      "grad_norm": 0.369140625,
      "learning_rate": 1.5060722676889516e-06,
      "loss": 0.7837,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6182
    },
    {
      "epoch": 0.840766929562143,
      "grad_norm": 0.42578125,
      "learning_rate": 1.5035675415497064e-06,
      "loss": 0.745,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6183
    },
    {
      "epoch": 0.8409029099809627,
      "grad_norm": 0.5859375,
      "learning_rate": 1.5010647306102121e-06,
      "loss": 0.84,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6184
    },
    {
      "epoch": 0.8410388903997824,
      "grad_norm": 0.30078125,
      "learning_rate": 1.4985638354346377e-06,
      "loss": 0.7352,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6185
    },
    {
      "epoch": 0.8411748708186021,
      "grad_norm": 0.5390625,
      "learning_rate": 1.496064856586713e-06,
      "loss": 0.7764,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6186
    },
    {
      "epoch": 0.8413108512374218,
      "grad_norm": 0.33984375,
      "learning_rate": 1.493567794629741e-06,
      "loss": 0.7004,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6187
    },
    {
      "epoch": 0.8414468316562415,
      "grad_norm": 0.44921875,
      "learning_rate": 1.491072650126587e-06,
      "loss": 0.4823,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6188
    },
    {
      "epoch": 0.8415828120750611,
      "grad_norm": 0.349609375,
      "learning_rate": 1.48857942363969e-06,
      "loss": 0.7192,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6189
    },
    {
      "epoch": 0.8417187924938809,
      "grad_norm": 0.369140625,
      "learning_rate": 1.4860881157310502e-06,
      "loss": 0.6852,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6190
    },
    {
      "epoch": 0.8418547729127006,
      "grad_norm": 0.5390625,
      "learning_rate": 1.483598726962243e-06,
      "loss": 0.8665,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6191
    },
    {
      "epoch": 0.8419907533315203,
      "grad_norm": 0.439453125,
      "learning_rate": 1.481111257894402e-06,
      "loss": 0.7516,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6192
    },
    {
      "epoch": 0.84212673375034,
      "grad_norm": 0.361328125,
      "learning_rate": 1.4786257090882393e-06,
      "loss": 0.627,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6193
    },
    {
      "epoch": 0.8422627141691597,
      "grad_norm": 0.484375,
      "learning_rate": 1.4761420811040229e-06,
      "loss": 0.7306,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6194
    },
    {
      "epoch": 0.8423986945879793,
      "grad_norm": 0.326171875,
      "learning_rate": 1.4736603745015964e-06,
      "loss": 0.5382,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6195
    },
    {
      "epoch": 0.842534675006799,
      "grad_norm": 0.30859375,
      "learning_rate": 1.471180589840363e-06,
      "loss": 0.6465,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6196
    },
    {
      "epoch": 0.8426706554256187,
      "grad_norm": 0.349609375,
      "learning_rate": 1.4687027276793008e-06,
      "loss": 0.4896,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6197
    },
    {
      "epoch": 0.8428066358444384,
      "grad_norm": 0.69140625,
      "learning_rate": 1.4662267885769455e-06,
      "loss": 0.7409,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6198
    },
    {
      "epoch": 0.8429426162632581,
      "grad_norm": 1.0546875,
      "learning_rate": 1.4637527730914092e-06,
      "loss": 0.6186,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6199
    },
    {
      "epoch": 0.8430785966820777,
      "grad_norm": 0.287109375,
      "learning_rate": 1.4612806817803604e-06,
      "loss": 0.5526,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6200
    },
    {
      "epoch": 0.8432145771008974,
      "grad_norm": 0.333984375,
      "learning_rate": 1.4588105152010435e-06,
      "loss": 0.6068,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6201
    },
    {
      "epoch": 0.8433505575197172,
      "grad_norm": 0.439453125,
      "learning_rate": 1.4563422739102596e-06,
      "loss": 0.7653,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6202
    },
    {
      "epoch": 0.8434865379385369,
      "grad_norm": 0.333984375,
      "learning_rate": 1.4538759584643847e-06,
      "loss": 0.498,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6203
    },
    {
      "epoch": 0.8436225183573566,
      "grad_norm": 0.392578125,
      "learning_rate": 1.451411569419352e-06,
      "loss": 0.7033,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6204
    },
    {
      "epoch": 0.8437584987761763,
      "grad_norm": 0.279296875,
      "learning_rate": 1.4489491073306693e-06,
      "loss": 0.5431,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6205
    },
    {
      "epoch": 0.8438944791949959,
      "grad_norm": 0.388671875,
      "learning_rate": 1.4464885727534018e-06,
      "loss": 0.5775,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6206
    },
    {
      "epoch": 0.8440304596138156,
      "grad_norm": 0.431640625,
      "learning_rate": 1.4440299662421887e-06,
      "loss": 0.652,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6207
    },
    {
      "epoch": 0.8441664400326353,
      "grad_norm": 0.609375,
      "learning_rate": 1.4415732883512235e-06,
      "loss": 0.6873,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6208
    },
    {
      "epoch": 0.844302420451455,
      "grad_norm": 0.427734375,
      "learning_rate": 1.4391185396342787e-06,
      "loss": 0.7572,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6209
    },
    {
      "epoch": 0.8444384008702747,
      "grad_norm": 0.318359375,
      "learning_rate": 1.4366657206446777e-06,
      "loss": 0.533,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6210
    },
    {
      "epoch": 0.8445743812890943,
      "grad_norm": 0.62109375,
      "learning_rate": 1.4342148319353222e-06,
      "loss": 0.7599,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6211
    },
    {
      "epoch": 0.844710361707914,
      "grad_norm": 0.443359375,
      "learning_rate": 1.431765874058667e-06,
      "loss": 0.7847,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6212
    },
    {
      "epoch": 0.8448463421267337,
      "grad_norm": 1.0390625,
      "learning_rate": 1.4293188475667418e-06,
      "loss": 0.7811,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6213
    },
    {
      "epoch": 0.8449823225455534,
      "grad_norm": 0.2734375,
      "learning_rate": 1.4268737530111332e-06,
      "loss": 0.4862,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6214
    },
    {
      "epoch": 0.8451183029643732,
      "grad_norm": 1.484375,
      "learning_rate": 1.4244305909429968e-06,
      "loss": 0.7189,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6215
    },
    {
      "epoch": 0.8452542833831929,
      "grad_norm": 0.380859375,
      "learning_rate": 1.4219893619130498e-06,
      "loss": 0.7246,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6216
    },
    {
      "epoch": 0.8453902638020125,
      "grad_norm": 0.330078125,
      "learning_rate": 1.4195500664715778e-06,
      "loss": 0.618,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6217
    },
    {
      "epoch": 0.8455262442208322,
      "grad_norm": 0.26953125,
      "learning_rate": 1.4171127051684242e-06,
      "loss": 0.5355,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6218
    },
    {
      "epoch": 0.8456622246396519,
      "grad_norm": 0.26953125,
      "learning_rate": 1.4146772785530038e-06,
      "loss": 0.4595,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6219
    },
    {
      "epoch": 0.8457982050584716,
      "grad_norm": 0.357421875,
      "learning_rate": 1.4122437871742877e-06,
      "loss": 0.7438,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6220
    },
    {
      "epoch": 0.8459341854772913,
      "grad_norm": 0.306640625,
      "learning_rate": 1.4098122315808194e-06,
      "loss": 0.4723,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6221
    },
    {
      "epoch": 0.846070165896111,
      "grad_norm": 0.359375,
      "learning_rate": 1.4073826123206946e-06,
      "loss": 0.7096,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6222
    },
    {
      "epoch": 0.8462061463149306,
      "grad_norm": 0.5234375,
      "learning_rate": 1.4049549299415866e-06,
      "loss": 0.5817,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6223
    },
    {
      "epoch": 0.8463421267337503,
      "grad_norm": 0.625,
      "learning_rate": 1.4025291849907186e-06,
      "loss": 0.5838,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6224
    },
    {
      "epoch": 0.84647810715257,
      "grad_norm": 0.26953125,
      "learning_rate": 1.4001053780148876e-06,
      "loss": 0.575,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6225
    },
    {
      "epoch": 0.8466140875713897,
      "grad_norm": 0.365234375,
      "learning_rate": 1.3976835095604446e-06,
      "loss": 0.6719,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6226
    },
    {
      "epoch": 0.8467500679902095,
      "grad_norm": 0.59765625,
      "learning_rate": 1.3952635801733138e-06,
      "loss": 0.6147,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6227
    },
    {
      "epoch": 0.8468860484090291,
      "grad_norm": 0.380859375,
      "learning_rate": 1.3928455903989724e-06,
      "loss": 0.7093,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6228
    },
    {
      "epoch": 0.8470220288278488,
      "grad_norm": 0.294921875,
      "learning_rate": 1.3904295407824686e-06,
      "loss": 0.5487,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6229
    },
    {
      "epoch": 0.8471580092466685,
      "grad_norm": 0.2177734375,
      "learning_rate": 1.388015431868408e-06,
      "loss": 0.3693,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6230
    },
    {
      "epoch": 0.8472939896654882,
      "grad_norm": 0.337890625,
      "learning_rate": 1.3856032642009576e-06,
      "loss": 0.598,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6231
    },
    {
      "epoch": 0.8474299700843079,
      "grad_norm": 0.30859375,
      "learning_rate": 1.3831930383238534e-06,
      "loss": 0.4945,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6232
    },
    {
      "epoch": 0.8475659505031276,
      "grad_norm": 0.4296875,
      "learning_rate": 1.3807847547803866e-06,
      "loss": 0.6198,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6233
    },
    {
      "epoch": 0.8477019309219472,
      "grad_norm": 0.375,
      "learning_rate": 1.3783784141134171e-06,
      "loss": 0.7093,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6234
    },
    {
      "epoch": 0.8478379113407669,
      "grad_norm": 0.423828125,
      "learning_rate": 1.375974016865359e-06,
      "loss": 0.5086,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6235
    },
    {
      "epoch": 0.8479738917595866,
      "grad_norm": 0.24609375,
      "learning_rate": 1.3735715635781988e-06,
      "loss": 0.3799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6236
    },
    {
      "epoch": 0.8481098721784063,
      "grad_norm": 0.314453125,
      "learning_rate": 1.371171054793473e-06,
      "loss": 0.5573,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6237
    },
    {
      "epoch": 0.848245852597226,
      "grad_norm": 0.68359375,
      "learning_rate": 1.3687724910522903e-06,
      "loss": 0.4863,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6238
    },
    {
      "epoch": 0.8483818330160456,
      "grad_norm": 0.462890625,
      "learning_rate": 1.3663758728953114e-06,
      "loss": 0.6906,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6239
    },
    {
      "epoch": 0.8485178134348654,
      "grad_norm": 0.33203125,
      "learning_rate": 1.3639812008627685e-06,
      "loss": 0.612,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6240
    },
    {
      "epoch": 0.8486537938536851,
      "grad_norm": 0.47265625,
      "learning_rate": 1.361588475494444e-06,
      "loss": 0.6073,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6241
    },
    {
      "epoch": 0.8487897742725048,
      "grad_norm": 0.357421875,
      "learning_rate": 1.359197697329695e-06,
      "loss": 0.6045,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6242
    },
    {
      "epoch": 0.8489257546913245,
      "grad_norm": 0.58203125,
      "learning_rate": 1.3568088669074254e-06,
      "loss": 0.6816,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6243
    },
    {
      "epoch": 0.8490617351101442,
      "grad_norm": 0.380859375,
      "learning_rate": 1.354421984766112e-06,
      "loss": 0.5245,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6244
    },
    {
      "epoch": 0.8491977155289638,
      "grad_norm": 0.24609375,
      "learning_rate": 1.352037051443783e-06,
      "loss": 0.451,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6245
    },
    {
      "epoch": 0.8493336959477835,
      "grad_norm": 0.4296875,
      "learning_rate": 1.3496540674780356e-06,
      "loss": 0.516,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6246
    },
    {
      "epoch": 0.8494696763666032,
      "grad_norm": 0.486328125,
      "learning_rate": 1.3472730334060192e-06,
      "loss": 0.744,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6247
    },
    {
      "epoch": 0.8496056567854229,
      "grad_norm": 0.6640625,
      "learning_rate": 1.3448939497644508e-06,
      "loss": 0.764,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6248
    },
    {
      "epoch": 0.8497416372042426,
      "grad_norm": 0.419921875,
      "learning_rate": 1.342516817089603e-06,
      "loss": 0.5986,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6249
    },
    {
      "epoch": 0.8498776176230622,
      "grad_norm": 0.412109375,
      "learning_rate": 1.3401416359173136e-06,
      "loss": 0.6278,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6250
    },
    {
      "epoch": 0.8500135980418819,
      "grad_norm": 0.25390625,
      "learning_rate": 1.3377684067829745e-06,
      "loss": 0.423,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6251
    },
    {
      "epoch": 0.8501495784607016,
      "grad_norm": 0.302734375,
      "learning_rate": 1.3353971302215418e-06,
      "loss": 0.6159,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6252
    },
    {
      "epoch": 0.8502855588795214,
      "grad_norm": 0.294921875,
      "learning_rate": 1.3330278067675294e-06,
      "loss": 0.6457,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6253
    },
    {
      "epoch": 0.8504215392983411,
      "grad_norm": 0.365234375,
      "learning_rate": 1.3306604369550147e-06,
      "loss": 0.585,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6254
    },
    {
      "epoch": 0.8505575197171608,
      "grad_norm": 0.443359375,
      "learning_rate": 1.328295021317626e-06,
      "loss": 0.7257,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6255
    },
    {
      "epoch": 0.8506935001359804,
      "grad_norm": 0.6015625,
      "learning_rate": 1.3259315603885637e-06,
      "loss": 0.9342,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6256
    },
    {
      "epoch": 0.8508294805548001,
      "grad_norm": 0.41015625,
      "learning_rate": 1.3235700547005758e-06,
      "loss": 0.6188,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6257
    },
    {
      "epoch": 0.8509654609736198,
      "grad_norm": 0.39453125,
      "learning_rate": 1.3212105047859768e-06,
      "loss": 0.5247,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6258
    },
    {
      "epoch": 0.8511014413924395,
      "grad_norm": 0.3203125,
      "learning_rate": 1.3188529111766368e-06,
      "loss": 0.5659,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6259
    },
    {
      "epoch": 0.8512374218112592,
      "grad_norm": 0.2333984375,
      "learning_rate": 1.3164972744039895e-06,
      "loss": 0.4569,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6260
    },
    {
      "epoch": 0.8513734022300788,
      "grad_norm": 0.421875,
      "learning_rate": 1.3141435949990188e-06,
      "loss": 0.5309,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6261
    },
    {
      "epoch": 0.8515093826488985,
      "grad_norm": 0.5546875,
      "learning_rate": 1.3117918734922775e-06,
      "loss": 0.6729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6262
    },
    {
      "epoch": 0.8516453630677182,
      "grad_norm": 0.400390625,
      "learning_rate": 1.3094421104138688e-06,
      "loss": 0.6943,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6263
    },
    {
      "epoch": 0.8517813434865379,
      "grad_norm": 0.515625,
      "learning_rate": 1.3070943062934627e-06,
      "loss": 0.784,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6264
    },
    {
      "epoch": 0.8519173239053577,
      "grad_norm": 0.333984375,
      "learning_rate": 1.304748461660278e-06,
      "loss": 0.5809,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6265
    },
    {
      "epoch": 0.8520533043241774,
      "grad_norm": 0.50390625,
      "learning_rate": 1.3024045770431004e-06,
      "loss": 0.8079,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6266
    },
    {
      "epoch": 0.852189284742997,
      "grad_norm": 0.361328125,
      "learning_rate": 1.3000626529702664e-06,
      "loss": 0.689,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6267
    },
    {
      "epoch": 0.8523252651618167,
      "grad_norm": 0.6328125,
      "learning_rate": 1.2977226899696781e-06,
      "loss": 0.8451,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6268
    },
    {
      "epoch": 0.8524612455806364,
      "grad_norm": 0.66015625,
      "learning_rate": 1.2953846885687893e-06,
      "loss": 0.682,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6269
    },
    {
      "epoch": 0.8525972259994561,
      "grad_norm": 0.51171875,
      "learning_rate": 1.293048649294617e-06,
      "loss": 0.6854,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6270
    },
    {
      "epoch": 0.8527332064182758,
      "grad_norm": 0.31640625,
      "learning_rate": 1.290714572673728e-06,
      "loss": 0.5073,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6271
    },
    {
      "epoch": 0.8528691868370954,
      "grad_norm": 0.341796875,
      "learning_rate": 1.288382459232258e-06,
      "loss": 0.6434,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6272
    },
    {
      "epoch": 0.8530051672559151,
      "grad_norm": 0.421875,
      "learning_rate": 1.2860523094958876e-06,
      "loss": 0.8908,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6273
    },
    {
      "epoch": 0.8531411476747348,
      "grad_norm": 0.287109375,
      "learning_rate": 1.2837241239898669e-06,
      "loss": 0.4588,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6274
    },
    {
      "epoch": 0.8532771280935545,
      "grad_norm": 0.6328125,
      "learning_rate": 1.2813979032389923e-06,
      "loss": 0.6112,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6275
    },
    {
      "epoch": 0.8534131085123742,
      "grad_norm": 0.6484375,
      "learning_rate": 1.2790736477676257e-06,
      "loss": 0.6396,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6276
    },
    {
      "epoch": 0.8535490889311939,
      "grad_norm": 0.5625,
      "learning_rate": 1.2767513580996805e-06,
      "loss": 0.6636,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6277
    },
    {
      "epoch": 0.8536850693500136,
      "grad_norm": 0.578125,
      "learning_rate": 1.2744310347586332e-06,
      "loss": 0.7764,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6278
    },
    {
      "epoch": 0.8538210497688333,
      "grad_norm": 0.5234375,
      "learning_rate": 1.2721126782675065e-06,
      "loss": 0.5841,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6279
    },
    {
      "epoch": 0.853957030187653,
      "grad_norm": 0.32421875,
      "learning_rate": 1.2697962891488924e-06,
      "loss": 0.4661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6280
    },
    {
      "epoch": 0.8540930106064727,
      "grad_norm": 0.25,
      "learning_rate": 1.267481867924929e-06,
      "loss": 0.388,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6281
    },
    {
      "epoch": 0.8542289910252924,
      "grad_norm": 0.44921875,
      "learning_rate": 1.2651694151173199e-06,
      "loss": 0.6984,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6282
    },
    {
      "epoch": 0.854364971444112,
      "grad_norm": 0.48046875,
      "learning_rate": 1.262858931247314e-06,
      "loss": 0.3963,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6283
    },
    {
      "epoch": 0.8545009518629317,
      "grad_norm": 0.328125,
      "learning_rate": 1.2605504168357286e-06,
      "loss": 0.5905,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6284
    },
    {
      "epoch": 0.8546369322817514,
      "grad_norm": 0.51953125,
      "learning_rate": 1.258243872402929e-06,
      "loss": 0.626,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6285
    },
    {
      "epoch": 0.8547729127005711,
      "grad_norm": 0.3046875,
      "learning_rate": 1.2559392984688357e-06,
      "loss": 0.4603,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6286
    },
    {
      "epoch": 0.8549088931193908,
      "grad_norm": 0.515625,
      "learning_rate": 1.253636695552931e-06,
      "loss": 0.6652,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6287
    },
    {
      "epoch": 0.8550448735382105,
      "grad_norm": 0.396484375,
      "learning_rate": 1.2513360641742477e-06,
      "loss": 0.7576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6288
    },
    {
      "epoch": 0.8551808539570301,
      "grad_norm": 0.43359375,
      "learning_rate": 1.2490374048513799e-06,
      "loss": 0.7227,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6289
    },
    {
      "epoch": 0.8553168343758499,
      "grad_norm": 0.3671875,
      "learning_rate": 1.2467407181024683e-06,
      "loss": 0.7466,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6290
    },
    {
      "epoch": 0.8554528147946696,
      "grad_norm": 0.388671875,
      "learning_rate": 1.2444460044452188e-06,
      "loss": 0.5451,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6291
    },
    {
      "epoch": 0.8555887952134893,
      "grad_norm": 0.78125,
      "learning_rate": 1.2421532643968847e-06,
      "loss": 0.9622,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6292
    },
    {
      "epoch": 0.855724775632309,
      "grad_norm": 0.318359375,
      "learning_rate": 1.2398624984742801e-06,
      "loss": 0.3983,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6293
    },
    {
      "epoch": 0.8558607560511287,
      "grad_norm": 0.54296875,
      "learning_rate": 1.2375737071937698e-06,
      "loss": 0.7711,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6294
    },
    {
      "epoch": 0.8559967364699483,
      "grad_norm": 0.58203125,
      "learning_rate": 1.2352868910712768e-06,
      "loss": 0.6139,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6295
    },
    {
      "epoch": 0.856132716888768,
      "grad_norm": 0.81640625,
      "learning_rate": 1.233002050622275e-06,
      "loss": 0.6727,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6296
    },
    {
      "epoch": 0.8562686973075877,
      "grad_norm": 0.60546875,
      "learning_rate": 1.2307191863617985e-06,
      "loss": 0.754,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6297
    },
    {
      "epoch": 0.8564046777264074,
      "grad_norm": 0.384765625,
      "learning_rate": 1.2284382988044297e-06,
      "loss": 0.708,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6298
    },
    {
      "epoch": 0.8565406581452271,
      "grad_norm": 0.302734375,
      "learning_rate": 1.226159388464313e-06,
      "loss": 0.6602,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6299
    },
    {
      "epoch": 0.8566766385640467,
      "grad_norm": 0.55859375,
      "learning_rate": 1.2238824558551365e-06,
      "loss": 0.5532,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6300
    },
    {
      "epoch": 0.8568126189828664,
      "grad_norm": 0.5234375,
      "learning_rate": 1.2216075014901529e-06,
      "loss": 0.568,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6301
    },
    {
      "epoch": 0.8569485994016861,
      "grad_norm": 0.55859375,
      "learning_rate": 1.2193345258821631e-06,
      "loss": 0.646,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6302
    },
    {
      "epoch": 0.8570845798205059,
      "grad_norm": 0.259765625,
      "learning_rate": 1.2170635295435263e-06,
      "loss": 0.5226,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6303
    },
    {
      "epoch": 0.8572205602393256,
      "grad_norm": 0.4765625,
      "learning_rate": 1.2147945129861494e-06,
      "loss": 0.5701,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6304
    },
    {
      "epoch": 0.8573565406581453,
      "grad_norm": 0.734375,
      "learning_rate": 1.2125274767215e-06,
      "loss": 0.5944,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6305
    },
    {
      "epoch": 0.8574925210769649,
      "grad_norm": 0.42578125,
      "learning_rate": 1.2102624212605908e-06,
      "loss": 0.8024,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6306
    },
    {
      "epoch": 0.8576285014957846,
      "grad_norm": 0.390625,
      "learning_rate": 1.2079993471139983e-06,
      "loss": 0.5143,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6307
    },
    {
      "epoch": 0.8577644819146043,
      "grad_norm": 0.484375,
      "learning_rate": 1.2057382547918428e-06,
      "loss": 0.7666,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6308
    },
    {
      "epoch": 0.857900462333424,
      "grad_norm": 0.46875,
      "learning_rate": 1.2034791448038063e-06,
      "loss": 0.5586,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6309
    },
    {
      "epoch": 0.8580364427522437,
      "grad_norm": 0.4453125,
      "learning_rate": 1.201222017659115e-06,
      "loss": 0.6574,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6310
    },
    {
      "epoch": 0.8581724231710633,
      "grad_norm": 0.34375,
      "learning_rate": 1.1989668738665582e-06,
      "loss": 0.6896,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6311
    },
    {
      "epoch": 0.858308403589883,
      "grad_norm": 0.67578125,
      "learning_rate": 1.1967137139344675e-06,
      "loss": 0.8574,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6312
    },
    {
      "epoch": 0.8584443840087027,
      "grad_norm": 0.412109375,
      "learning_rate": 1.1944625383707377e-06,
      "loss": 0.7106,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6313
    },
    {
      "epoch": 0.8585803644275224,
      "grad_norm": 0.306640625,
      "learning_rate": 1.1922133476828058e-06,
      "loss": 0.5368,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6314
    },
    {
      "epoch": 0.8587163448463421,
      "grad_norm": 0.3828125,
      "learning_rate": 1.1899661423776732e-06,
      "loss": 0.6864,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6315
    },
    {
      "epoch": 0.8588523252651619,
      "grad_norm": 0.435546875,
      "learning_rate": 1.1877209229618825e-06,
      "loss": 0.6187,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6316
    },
    {
      "epoch": 0.8589883056839815,
      "grad_norm": 0.412109375,
      "learning_rate": 1.1854776899415366e-06,
      "loss": 0.8335,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6317
    },
    {
      "epoch": 0.8591242861028012,
      "grad_norm": 0.62109375,
      "learning_rate": 1.1832364438222855e-06,
      "loss": 0.744,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6318
    },
    {
      "epoch": 0.8592602665216209,
      "grad_norm": 0.294921875,
      "learning_rate": 1.1809971851093349e-06,
      "loss": 0.5716,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6319
    },
    {
      "epoch": 0.8593962469404406,
      "grad_norm": 0.443359375,
      "learning_rate": 1.178759914307439e-06,
      "loss": 0.7664,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6320
    },
    {
      "epoch": 0.8595322273592603,
      "grad_norm": 0.34765625,
      "learning_rate": 1.1765246319209111e-06,
      "loss": 0.6208,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6321
    },
    {
      "epoch": 0.85966820777808,
      "grad_norm": 0.412109375,
      "learning_rate": 1.1742913384536036e-06,
      "loss": 0.7773,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6322
    },
    {
      "epoch": 0.8598041881968996,
      "grad_norm": 0.455078125,
      "learning_rate": 1.1720600344089361e-06,
      "loss": 0.6089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6323
    },
    {
      "epoch": 0.8599401686157193,
      "grad_norm": 0.28125,
      "learning_rate": 1.1698307202898651e-06,
      "loss": 0.5256,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6324
    },
    {
      "epoch": 0.860076149034539,
      "grad_norm": 0.392578125,
      "learning_rate": 1.1676033965989098e-06,
      "loss": 0.7031,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6325
    },
    {
      "epoch": 0.8602121294533587,
      "grad_norm": 0.34375,
      "learning_rate": 1.165378063838133e-06,
      "loss": 0.4371,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6326
    },
    {
      "epoch": 0.8603481098721784,
      "grad_norm": 0.451171875,
      "learning_rate": 1.1631547225091545e-06,
      "loss": 0.7064,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6327
    },
    {
      "epoch": 0.8604840902909981,
      "grad_norm": 0.37109375,
      "learning_rate": 1.1609333731131401e-06,
      "loss": 0.6099,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6328
    },
    {
      "epoch": 0.8606200707098178,
      "grad_norm": 0.328125,
      "learning_rate": 1.1587140161508126e-06,
      "loss": 0.5697,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6329
    },
    {
      "epoch": 0.8607560511286375,
      "grad_norm": 0.2412109375,
      "learning_rate": 1.1564966521224386e-06,
      "loss": 0.4137,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6330
    },
    {
      "epoch": 0.8608920315474572,
      "grad_norm": 0.357421875,
      "learning_rate": 1.1542812815278414e-06,
      "loss": 0.5441,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6331
    },
    {
      "epoch": 0.8610280119662769,
      "grad_norm": 0.240234375,
      "learning_rate": 1.1520679048663908e-06,
      "loss": 0.467,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6332
    },
    {
      "epoch": 0.8611639923850966,
      "grad_norm": 0.29296875,
      "learning_rate": 1.1498565226370118e-06,
      "loss": 0.519,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6333
    },
    {
      "epoch": 0.8612999728039162,
      "grad_norm": 0.5078125,
      "learning_rate": 1.1476471353381735e-06,
      "loss": 0.7173,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6334
    },
    {
      "epoch": 0.8614359532227359,
      "grad_norm": 0.318359375,
      "learning_rate": 1.1454397434679022e-06,
      "loss": 0.7349,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6335
    },
    {
      "epoch": 0.8615719336415556,
      "grad_norm": 0.453125,
      "learning_rate": 1.143234347523766e-06,
      "loss": 0.6773,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6336
    },
    {
      "epoch": 0.8617079140603753,
      "grad_norm": 0.26953125,
      "learning_rate": 1.1410309480028947e-06,
      "loss": 0.4331,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6337
    },
    {
      "epoch": 0.861843894479195,
      "grad_norm": 0.3671875,
      "learning_rate": 1.138829545401956e-06,
      "loss": 0.6681,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6338
    },
    {
      "epoch": 0.8619798748980146,
      "grad_norm": 0.71875,
      "learning_rate": 1.1366301402171775e-06,
      "loss": 0.8851,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6339
    },
    {
      "epoch": 0.8621158553168343,
      "grad_norm": 0.341796875,
      "learning_rate": 1.1344327329443293e-06,
      "loss": 0.4831,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6340
    },
    {
      "epoch": 0.8622518357356541,
      "grad_norm": 0.291015625,
      "learning_rate": 1.1322373240787321e-06,
      "loss": 0.6061,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6341
    },
    {
      "epoch": 0.8623878161544738,
      "grad_norm": 0.287109375,
      "learning_rate": 1.130043914115262e-06,
      "loss": 0.603,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6342
    },
    {
      "epoch": 0.8625237965732935,
      "grad_norm": 0.37109375,
      "learning_rate": 1.1278525035483367e-06,
      "loss": 0.5591,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6343
    },
    {
      "epoch": 0.8626597769921132,
      "grad_norm": 0.3125,
      "learning_rate": 1.1256630928719302e-06,
      "loss": 0.6956,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6344
    },
    {
      "epoch": 0.8627957574109328,
      "grad_norm": 0.859375,
      "learning_rate": 1.12347568257956e-06,
      "loss": 0.8364,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6345
    },
    {
      "epoch": 0.8629317378297525,
      "grad_norm": 0.404296875,
      "learning_rate": 1.1212902731642972e-06,
      "loss": 0.6657,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6346
    },
    {
      "epoch": 0.8630677182485722,
      "grad_norm": 0.318359375,
      "learning_rate": 1.1191068651187564e-06,
      "loss": 0.6702,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6347
    },
    {
      "epoch": 0.8632036986673919,
      "grad_norm": 0.427734375,
      "learning_rate": 1.1169254589351096e-06,
      "loss": 0.5622,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6348
    },
    {
      "epoch": 0.8633396790862116,
      "grad_norm": 0.46484375,
      "learning_rate": 1.114746055105067e-06,
      "loss": 0.6904,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6349
    },
    {
      "epoch": 0.8634756595050312,
      "grad_norm": 0.427734375,
      "learning_rate": 1.1125686541198966e-06,
      "loss": 0.6689,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6350
    },
    {
      "epoch": 0.8636116399238509,
      "grad_norm": 0.296875,
      "learning_rate": 1.110393256470409e-06,
      "loss": 0.6103,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6351
    },
    {
      "epoch": 0.8637476203426706,
      "grad_norm": 0.41796875,
      "learning_rate": 1.1082198626469687e-06,
      "loss": 0.8338,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6352
    },
    {
      "epoch": 0.8638836007614904,
      "grad_norm": 0.376953125,
      "learning_rate": 1.1060484731394805e-06,
      "loss": 0.6452,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6353
    },
    {
      "epoch": 0.8640195811803101,
      "grad_norm": 0.49609375,
      "learning_rate": 1.1038790884374062e-06,
      "loss": 0.4474,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6354
    },
    {
      "epoch": 0.8641555615991298,
      "grad_norm": 0.36328125,
      "learning_rate": 1.101711709029748e-06,
      "loss": 0.5776,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6355
    },
    {
      "epoch": 0.8642915420179494,
      "grad_norm": 0.4296875,
      "learning_rate": 1.099546335405065e-06,
      "loss": 0.7514,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6356
    },
    {
      "epoch": 0.8644275224367691,
      "grad_norm": 0.640625,
      "learning_rate": 1.0973829680514525e-06,
      "loss": 0.5153,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6357
    },
    {
      "epoch": 0.8645635028555888,
      "grad_norm": 0.80078125,
      "learning_rate": 1.0952216074565658e-06,
      "loss": 0.6471,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6358
    },
    {
      "epoch": 0.8646994832744085,
      "grad_norm": 0.2578125,
      "learning_rate": 1.0930622541075963e-06,
      "loss": 0.5361,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6359
    },
    {
      "epoch": 0.8648354636932282,
      "grad_norm": 1.140625,
      "learning_rate": 1.0909049084912948e-06,
      "loss": 0.7389,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6360
    },
    {
      "epoch": 0.8649714441120478,
      "grad_norm": 0.1845703125,
      "learning_rate": 1.0887495710939455e-06,
      "loss": 0.3566,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6361
    },
    {
      "epoch": 0.8651074245308675,
      "grad_norm": 0.26171875,
      "learning_rate": 1.0865962424013975e-06,
      "loss": 0.4657,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6362
    },
    {
      "epoch": 0.8652434049496872,
      "grad_norm": 0.498046875,
      "learning_rate": 1.0844449228990294e-06,
      "loss": 0.8089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6363
    },
    {
      "epoch": 0.8653793853685069,
      "grad_norm": 0.4609375,
      "learning_rate": 1.08229561307178e-06,
      "loss": 0.7152,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6364
    },
    {
      "epoch": 0.8655153657873266,
      "grad_norm": 0.3125,
      "learning_rate": 1.080148313404127e-06,
      "loss": 0.6457,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6365
    },
    {
      "epoch": 0.8656513462061464,
      "grad_norm": 0.328125,
      "learning_rate": 1.0780030243801e-06,
      "loss": 0.5814,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6366
    },
    {
      "epoch": 0.865787326624966,
      "grad_norm": 0.306640625,
      "learning_rate": 1.0758597464832709e-06,
      "loss": 0.5444,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6367
    },
    {
      "epoch": 0.8659233070437857,
      "grad_norm": 0.380859375,
      "learning_rate": 1.0737184801967637e-06,
      "loss": 0.7056,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6368
    },
    {
      "epoch": 0.8660592874626054,
      "grad_norm": 0.369140625,
      "learning_rate": 1.0715792260032431e-06,
      "loss": 0.7274,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6369
    },
    {
      "epoch": 0.8661952678814251,
      "grad_norm": 0.482421875,
      "learning_rate": 1.0694419843849258e-06,
      "loss": 0.6997,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6370
    },
    {
      "epoch": 0.8663312483002448,
      "grad_norm": 0.41796875,
      "learning_rate": 1.0673067558235695e-06,
      "loss": 0.6799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6371
    },
    {
      "epoch": 0.8664672287190645,
      "grad_norm": 0.435546875,
      "learning_rate": 1.0651735408004838e-06,
      "loss": 0.7671,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6372
    },
    {
      "epoch": 0.8666032091378841,
      "grad_norm": 0.640625,
      "learning_rate": 1.0630423397965184e-06,
      "loss": 0.616,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6373
    },
    {
      "epoch": 0.8667391895567038,
      "grad_norm": 0.3125,
      "learning_rate": 1.0609131532920746e-06,
      "loss": 0.6572,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6374
    },
    {
      "epoch": 0.8668751699755235,
      "grad_norm": 0.453125,
      "learning_rate": 1.0587859817670953e-06,
      "loss": 0.7669,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6375
    },
    {
      "epoch": 0.8670111503943432,
      "grad_norm": 0.357421875,
      "learning_rate": 1.0566608257010736e-06,
      "loss": 0.5745,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6376
    },
    {
      "epoch": 0.8671471308131629,
      "grad_norm": 0.28515625,
      "learning_rate": 1.0545376855730405e-06,
      "loss": 0.5376,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6377
    },
    {
      "epoch": 0.8672831112319825,
      "grad_norm": 0.486328125,
      "learning_rate": 1.0524165618615845e-06,
      "loss": 0.341,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6378
    },
    {
      "epoch": 0.8674190916508023,
      "grad_norm": 0.310546875,
      "learning_rate": 1.0502974550448264e-06,
      "loss": 0.7002,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6379
    },
    {
      "epoch": 0.867555072069622,
      "grad_norm": 0.435546875,
      "learning_rate": 1.048180365600444e-06,
      "loss": 0.5278,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6380
    },
    {
      "epoch": 0.8676910524884417,
      "grad_norm": 0.34765625,
      "learning_rate": 1.0460652940056505e-06,
      "loss": 0.5405,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6381
    },
    {
      "epoch": 0.8678270329072614,
      "grad_norm": 1.7109375,
      "learning_rate": 1.0439522407372137e-06,
      "loss": 0.8293,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6382
    },
    {
      "epoch": 0.8679630133260811,
      "grad_norm": 0.32421875,
      "learning_rate": 1.0418412062714367e-06,
      "loss": 0.5688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6383
    },
    {
      "epoch": 0.8680989937449007,
      "grad_norm": 0.2373046875,
      "learning_rate": 1.0397321910841772e-06,
      "loss": 0.5054,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6384
    },
    {
      "epoch": 0.8682349741637204,
      "grad_norm": 0.392578125,
      "learning_rate": 1.0376251956508299e-06,
      "loss": 0.6279,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6385
    },
    {
      "epoch": 0.8683709545825401,
      "grad_norm": 0.498046875,
      "learning_rate": 1.0355202204463389e-06,
      "loss": 0.8089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6386
    },
    {
      "epoch": 0.8685069350013598,
      "grad_norm": 0.48046875,
      "learning_rate": 1.033417265945189e-06,
      "loss": 0.7227,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6387
    },
    {
      "epoch": 0.8686429154201795,
      "grad_norm": 0.2890625,
      "learning_rate": 1.0313163326214149e-06,
      "loss": 0.5348,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6388
    },
    {
      "epoch": 0.8687788958389991,
      "grad_norm": 0.384765625,
      "learning_rate": 1.0292174209485906e-06,
      "loss": 0.6668,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6389
    },
    {
      "epoch": 0.8689148762578188,
      "grad_norm": 0.59765625,
      "learning_rate": 1.0271205313998378e-06,
      "loss": 0.6047,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6390
    },
    {
      "epoch": 0.8690508566766386,
      "grad_norm": 0.5078125,
      "learning_rate": 1.0250256644478196e-06,
      "loss": 0.7101,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6391
    },
    {
      "epoch": 0.8691868370954583,
      "grad_norm": 0.314453125,
      "learning_rate": 1.022932820564746e-06,
      "loss": 0.5591,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6392
    },
    {
      "epoch": 0.869322817514278,
      "grad_norm": 0.345703125,
      "learning_rate": 1.0208420002223696e-06,
      "loss": 0.5579,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6393
    },
    {
      "epoch": 0.8694587979330977,
      "grad_norm": 0.53125,
      "learning_rate": 1.0187532038919844e-06,
      "loss": 0.6669,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6394
    },
    {
      "epoch": 0.8695947783519173,
      "grad_norm": 0.388671875,
      "learning_rate": 1.0166664320444342e-06,
      "loss": 0.6511,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6395
    },
    {
      "epoch": 0.869730758770737,
      "grad_norm": 0.400390625,
      "learning_rate": 1.0145816851501001e-06,
      "loss": 0.5897,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6396
    },
    {
      "epoch": 0.8698667391895567,
      "grad_norm": 0.384765625,
      "learning_rate": 1.0124989636789117e-06,
      "loss": 0.4982,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6397
    },
    {
      "epoch": 0.8700027196083764,
      "grad_norm": 0.326171875,
      "learning_rate": 1.010418268100337e-06,
      "loss": 0.6051,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6398
    },
    {
      "epoch": 0.8701387000271961,
      "grad_norm": 0.3515625,
      "learning_rate": 1.0083395988833944e-06,
      "loss": 0.4961,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6399
    },
    {
      "epoch": 0.8702746804460157,
      "grad_norm": 0.28515625,
      "learning_rate": 1.0062629564966365e-06,
      "loss": 0.6541,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6400
    },
    {
      "epoch": 0.8704106608648354,
      "grad_norm": 0.55078125,
      "learning_rate": 1.0041883414081676e-06,
      "loss": 0.6585,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6401
    },
    {
      "epoch": 0.8705466412836551,
      "grad_norm": 0.251953125,
      "learning_rate": 1.0021157540856296e-06,
      "loss": 0.5163,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6402
    },
    {
      "epoch": 0.8706826217024748,
      "grad_norm": 0.259765625,
      "learning_rate": 1.000045194996211e-06,
      "loss": 0.5469,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6403
    },
    {
      "epoch": 0.8708186021212946,
      "grad_norm": 0.361328125,
      "learning_rate": 9.979766646066368e-07,
      "loss": 0.6309,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6404
    },
    {
      "epoch": 0.8709545825401143,
      "grad_norm": 0.48828125,
      "learning_rate": 9.959101633831848e-07,
      "loss": 0.6328,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6405
    },
    {
      "epoch": 0.871090562958934,
      "grad_norm": 0.326171875,
      "learning_rate": 9.938456917916638e-07,
      "loss": 0.6201,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6406
    },
    {
      "epoch": 0.8712265433777536,
      "grad_norm": 0.298828125,
      "learning_rate": 9.917832502974367e-07,
      "loss": 0.5957,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6407
    },
    {
      "epoch": 0.8713625237965733,
      "grad_norm": 0.37109375,
      "learning_rate": 9.89722839365398e-07,
      "loss": 0.589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6408
    },
    {
      "epoch": 0.871498504215393,
      "grad_norm": 0.34765625,
      "learning_rate": 9.876644594599927e-07,
      "loss": 0.6836,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6409
    },
    {
      "epoch": 0.8716344846342127,
      "grad_norm": 0.419921875,
      "learning_rate": 9.856081110452032e-07,
      "loss": 0.6019,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6410
    },
    {
      "epoch": 0.8717704650530324,
      "grad_norm": 0.2490234375,
      "learning_rate": 9.835537945845574e-07,
      "loss": 0.5866,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6411
    },
    {
      "epoch": 0.871906445471852,
      "grad_norm": 0.283203125,
      "learning_rate": 9.815015105411196e-07,
      "loss": 0.6379,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6412
    },
    {
      "epoch": 0.8720424258906717,
      "grad_norm": 0.353515625,
      "learning_rate": 9.794512593775052e-07,
      "loss": 0.5579,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6413
    },
    {
      "epoch": 0.8721784063094914,
      "grad_norm": 0.35546875,
      "learning_rate": 9.77403041555861e-07,
      "loss": 0.7823,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6414
    },
    {
      "epoch": 0.8723143867283111,
      "grad_norm": 0.2236328125,
      "learning_rate": 9.753568575378836e-07,
      "loss": 0.4347,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6415
    },
    {
      "epoch": 0.8724503671471309,
      "grad_norm": 0.349609375,
      "learning_rate": 9.733127077848047e-07,
      "loss": 0.688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6416
    },
    {
      "epoch": 0.8725863475659505,
      "grad_norm": 0.3671875,
      "learning_rate": 9.71270592757404e-07,
      "loss": 0.5841,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6417
    },
    {
      "epoch": 0.8727223279847702,
      "grad_norm": 0.421875,
      "learning_rate": 9.692305129159962e-07,
      "loss": 0.612,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6418
    },
    {
      "epoch": 0.8728583084035899,
      "grad_norm": 0.291015625,
      "learning_rate": 9.671924687204437e-07,
      "loss": 0.5428,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6419
    },
    {
      "epoch": 0.8729942888224096,
      "grad_norm": 0.55078125,
      "learning_rate": 9.651564606301422e-07,
      "loss": 0.6032,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6420
    },
    {
      "epoch": 0.8731302692412293,
      "grad_norm": 0.408203125,
      "learning_rate": 9.631224891040348e-07,
      "loss": 0.6322,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6421
    },
    {
      "epoch": 0.873266249660049,
      "grad_norm": 0.55859375,
      "learning_rate": 9.610905546006055e-07,
      "loss": 0.641,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6422
    },
    {
      "epoch": 0.8734022300788686,
      "grad_norm": 0.6171875,
      "learning_rate": 9.590606575778761e-07,
      "loss": 0.8172,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6423
    },
    {
      "epoch": 0.8735382104976883,
      "grad_norm": 0.376953125,
      "learning_rate": 9.57032798493408e-07,
      "loss": 0.6631,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6424
    },
    {
      "epoch": 0.873674190916508,
      "grad_norm": 0.76171875,
      "learning_rate": 9.550069778043103e-07,
      "loss": 0.6882,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6425
    },
    {
      "epoch": 0.8738101713353277,
      "grad_norm": 0.39453125,
      "learning_rate": 9.529831959672209e-07,
      "loss": 0.7307,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6426
    },
    {
      "epoch": 0.8739461517541474,
      "grad_norm": 0.4140625,
      "learning_rate": 9.509614534383315e-07,
      "loss": 0.6909,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6427
    },
    {
      "epoch": 0.874082132172967,
      "grad_norm": 0.71484375,
      "learning_rate": 9.489417506733633e-07,
      "loss": 0.5623,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6428
    },
    {
      "epoch": 0.8742181125917868,
      "grad_norm": 0.6171875,
      "learning_rate": 9.469240881275842e-07,
      "loss": 0.7581,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6429
    },
    {
      "epoch": 0.8743540930106065,
      "grad_norm": 0.404296875,
      "learning_rate": 9.449084662557984e-07,
      "loss": 0.7238,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6430
    },
    {
      "epoch": 0.8744900734294262,
      "grad_norm": 0.65625,
      "learning_rate": 9.428948855123543e-07,
      "loss": 0.596,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6431
    },
    {
      "epoch": 0.8746260538482459,
      "grad_norm": 0.2314453125,
      "learning_rate": 9.408833463511335e-07,
      "loss": 0.4661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6432
    },
    {
      "epoch": 0.8747620342670656,
      "grad_norm": 0.38671875,
      "learning_rate": 9.388738492255667e-07,
      "loss": 0.5924,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6433
    },
    {
      "epoch": 0.8748980146858852,
      "grad_norm": 0.296875,
      "learning_rate": 9.368663945886136e-07,
      "loss": 0.5933,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6434
    },
    {
      "epoch": 0.8750339951047049,
      "grad_norm": 0.30859375,
      "learning_rate": 9.348609828927835e-07,
      "loss": 0.6898,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6435
    },
    {
      "epoch": 0.8751699755235246,
      "grad_norm": 0.515625,
      "learning_rate": 9.32857614590118e-07,
      "loss": 0.6414,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6436
    },
    {
      "epoch": 0.8753059559423443,
      "grad_norm": 0.36328125,
      "learning_rate": 9.308562901322027e-07,
      "loss": 0.7266,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6437
    },
    {
      "epoch": 0.875441936361164,
      "grad_norm": 0.51171875,
      "learning_rate": 9.288570099701588e-07,
      "loss": 0.8396,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6438
    },
    {
      "epoch": 0.8755779167799836,
      "grad_norm": 0.291015625,
      "learning_rate": 9.268597745546504e-07,
      "loss": 0.5653,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6439
    },
    {
      "epoch": 0.8757138971988033,
      "grad_norm": 0.439453125,
      "learning_rate": 9.248645843358761e-07,
      "loss": 0.6792,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6440
    },
    {
      "epoch": 0.875849877617623,
      "grad_norm": 0.34375,
      "learning_rate": 9.228714397635796e-07,
      "loss": 0.6427,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6441
    },
    {
      "epoch": 0.8759858580364428,
      "grad_norm": 0.322265625,
      "learning_rate": 9.208803412870371e-07,
      "loss": 0.6711,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6442
    },
    {
      "epoch": 0.8761218384552625,
      "grad_norm": 0.3359375,
      "learning_rate": 9.188912893550694e-07,
      "loss": 0.6089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6443
    },
    {
      "epoch": 0.8762578188740822,
      "grad_norm": 0.88671875,
      "learning_rate": 9.169042844160292e-07,
      "loss": 0.7713,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6444
    },
    {
      "epoch": 0.8763937992929018,
      "grad_norm": 0.46875,
      "learning_rate": 9.149193269178158e-07,
      "loss": 0.6034,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6445
    },
    {
      "epoch": 0.8765297797117215,
      "grad_norm": 0.21484375,
      "learning_rate": 9.129364173078593e-07,
      "loss": 0.4155,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6446
    },
    {
      "epoch": 0.8766657601305412,
      "grad_norm": 0.30859375,
      "learning_rate": 9.109555560331352e-07,
      "loss": 0.5202,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6447
    },
    {
      "epoch": 0.8768017405493609,
      "grad_norm": 0.357421875,
      "learning_rate": 9.08976743540152e-07,
      "loss": 0.5946,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6448
    },
    {
      "epoch": 0.8769377209681806,
      "grad_norm": 0.302734375,
      "learning_rate": 9.069999802749563e-07,
      "loss": 0.5798,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6449
    },
    {
      "epoch": 0.8770737013870002,
      "grad_norm": 0.298828125,
      "learning_rate": 9.050252666831394e-07,
      "loss": 0.5526,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6450
    },
    {
      "epoch": 0.8772096818058199,
      "grad_norm": 0.322265625,
      "learning_rate": 9.030526032098208e-07,
      "loss": 0.5551,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6451
    },
    {
      "epoch": 0.8773456622246396,
      "grad_norm": 0.36328125,
      "learning_rate": 9.010819902996681e-07,
      "loss": 0.5117,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6452
    },
    {
      "epoch": 0.8774816426434593,
      "grad_norm": 0.2333984375,
      "learning_rate": 8.991134283968761e-07,
      "loss": 0.4784,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6453
    },
    {
      "epoch": 0.8776176230622791,
      "grad_norm": 0.34765625,
      "learning_rate": 8.971469179451886e-07,
      "loss": 0.5387,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6454
    },
    {
      "epoch": 0.8777536034810988,
      "grad_norm": 0.58984375,
      "learning_rate": 8.951824593878744e-07,
      "loss": 0.5085,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6455
    },
    {
      "epoch": 0.8778895838999184,
      "grad_norm": 0.37109375,
      "learning_rate": 8.932200531677537e-07,
      "loss": 0.5903,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6456
    },
    {
      "epoch": 0.8780255643187381,
      "grad_norm": 0.32421875,
      "learning_rate": 8.912596997271706e-07,
      "loss": 0.5674,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6457
    },
    {
      "epoch": 0.8781615447375578,
      "grad_norm": 0.32421875,
      "learning_rate": 8.893013995080169e-07,
      "loss": 0.5786,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6458
    },
    {
      "epoch": 0.8782975251563775,
      "grad_norm": 0.71484375,
      "learning_rate": 8.873451529517141e-07,
      "loss": 0.8223,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6459
    },
    {
      "epoch": 0.8784335055751972,
      "grad_norm": 0.3671875,
      "learning_rate": 8.853909604992284e-07,
      "loss": 0.487,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6460
    },
    {
      "epoch": 0.8785694859940169,
      "grad_norm": 0.455078125,
      "learning_rate": 8.834388225910528e-07,
      "loss": 0.8341,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6461
    },
    {
      "epoch": 0.8787054664128365,
      "grad_norm": 0.294921875,
      "learning_rate": 8.814887396672301e-07,
      "loss": 0.6137,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6462
    },
    {
      "epoch": 0.8788414468316562,
      "grad_norm": 0.60546875,
      "learning_rate": 8.795407121673261e-07,
      "loss": 0.765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6463
    },
    {
      "epoch": 0.8789774272504759,
      "grad_norm": 0.734375,
      "learning_rate": 8.775947405304553e-07,
      "loss": 0.6405,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6464
    },
    {
      "epoch": 0.8791134076692956,
      "grad_norm": 0.39453125,
      "learning_rate": 8.756508251952589e-07,
      "loss": 0.8468,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6465
    },
    {
      "epoch": 0.8792493880881153,
      "grad_norm": 0.6171875,
      "learning_rate": 8.737089665999243e-07,
      "loss": 0.4619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6466
    },
    {
      "epoch": 0.879385368506935,
      "grad_norm": 0.279296875,
      "learning_rate": 8.717691651821647e-07,
      "loss": 0.4056,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6467
    },
    {
      "epoch": 0.8795213489257547,
      "grad_norm": 0.451171875,
      "learning_rate": 8.698314213792402e-07,
      "loss": 0.7489,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6468
    },
    {
      "epoch": 0.8796573293445744,
      "grad_norm": 0.388671875,
      "learning_rate": 8.678957356279372e-07,
      "loss": 0.7773,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6469
    },
    {
      "epoch": 0.8797933097633941,
      "grad_norm": 0.31640625,
      "learning_rate": 8.659621083645875e-07,
      "loss": 0.613,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6470
    },
    {
      "epoch": 0.8799292901822138,
      "grad_norm": 0.287109375,
      "learning_rate": 8.640305400250492e-07,
      "loss": 0.6097,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6471
    },
    {
      "epoch": 0.8800652706010335,
      "grad_norm": 0.52734375,
      "learning_rate": 8.621010310447276e-07,
      "loss": 0.5123,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6472
    },
    {
      "epoch": 0.8802012510198531,
      "grad_norm": 0.357421875,
      "learning_rate": 8.60173581858551e-07,
      "loss": 0.7135,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6473
    },
    {
      "epoch": 0.8803372314386728,
      "grad_norm": 0.46484375,
      "learning_rate": 8.582481929009956e-07,
      "loss": 0.6868,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6474
    },
    {
      "epoch": 0.8804732118574925,
      "grad_norm": 0.28515625,
      "learning_rate": 8.56324864606064e-07,
      "loss": 0.4907,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6475
    },
    {
      "epoch": 0.8806091922763122,
      "grad_norm": 0.302734375,
      "learning_rate": 8.544035974073006e-07,
      "loss": 0.542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6476
    },
    {
      "epoch": 0.8807451726951319,
      "grad_norm": 0.423828125,
      "learning_rate": 8.524843917377801e-07,
      "loss": 0.5659,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6477
    },
    {
      "epoch": 0.8808811531139515,
      "grad_norm": 0.80859375,
      "learning_rate": 8.505672480301164e-07,
      "loss": 0.7273,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6478
    },
    {
      "epoch": 0.8810171335327713,
      "grad_norm": 0.451171875,
      "learning_rate": 8.48652166716456e-07,
      "loss": 0.7965,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6479
    },
    {
      "epoch": 0.881153113951591,
      "grad_norm": 0.478515625,
      "learning_rate": 8.467391482284836e-07,
      "loss": 0.7038,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6480
    },
    {
      "epoch": 0.8812890943704107,
      "grad_norm": 0.625,
      "learning_rate": 8.448281929974122e-07,
      "loss": 0.729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6481
    },
    {
      "epoch": 0.8814250747892304,
      "grad_norm": 0.4296875,
      "learning_rate": 8.429193014540015e-07,
      "loss": 0.8701,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6482
    },
    {
      "epoch": 0.8815610552080501,
      "grad_norm": 0.341796875,
      "learning_rate": 8.410124740285341e-07,
      "loss": 0.568,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6483
    },
    {
      "epoch": 0.8816970356268697,
      "grad_norm": 0.392578125,
      "learning_rate": 8.39107711150835e-07,
      "loss": 0.7746,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6484
    },
    {
      "epoch": 0.8818330160456894,
      "grad_norm": 0.416015625,
      "learning_rate": 8.372050132502573e-07,
      "loss": 0.5785,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6485
    },
    {
      "epoch": 0.8819689964645091,
      "grad_norm": 0.51171875,
      "learning_rate": 8.35304380755697e-07,
      "loss": 0.8867,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6486
    },
    {
      "epoch": 0.8821049768833288,
      "grad_norm": 0.578125,
      "learning_rate": 8.334058140955759e-07,
      "loss": 0.799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6487
    },
    {
      "epoch": 0.8822409573021485,
      "grad_norm": 0.294921875,
      "learning_rate": 8.315093136978569e-07,
      "loss": 0.5908,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6488
    },
    {
      "epoch": 0.8823769377209681,
      "grad_norm": 0.453125,
      "learning_rate": 8.296148799900328e-07,
      "loss": 0.3218,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6489
    },
    {
      "epoch": 0.8825129181397878,
      "grad_norm": 0.279296875,
      "learning_rate": 8.277225133991329e-07,
      "loss": 0.5257,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6490
    },
    {
      "epoch": 0.8826488985586075,
      "grad_norm": 0.330078125,
      "learning_rate": 8.258322143517183e-07,
      "loss": 0.5216,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6491
    },
    {
      "epoch": 0.8827848789774273,
      "grad_norm": 0.39453125,
      "learning_rate": 8.239439832738893e-07,
      "loss": 0.646,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6492
    },
    {
      "epoch": 0.882920859396247,
      "grad_norm": 0.51953125,
      "learning_rate": 8.220578205912711e-07,
      "loss": 0.7891,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6493
    },
    {
      "epoch": 0.8830568398150667,
      "grad_norm": 0.37109375,
      "learning_rate": 8.201737267290321e-07,
      "loss": 0.6981,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6494
    },
    {
      "epoch": 0.8831928202338863,
      "grad_norm": 0.33203125,
      "learning_rate": 8.182917021118664e-07,
      "loss": 0.5876,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6495
    },
    {
      "epoch": 0.883328800652706,
      "grad_norm": 0.494140625,
      "learning_rate": 8.164117471640098e-07,
      "loss": 0.7142,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6496
    },
    {
      "epoch": 0.8834647810715257,
      "grad_norm": 0.375,
      "learning_rate": 8.145338623092225e-07,
      "loss": 0.7745,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6497
    },
    {
      "epoch": 0.8836007614903454,
      "grad_norm": 0.4453125,
      "learning_rate": 8.126580479708068e-07,
      "loss": 0.7193,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6498
    },
    {
      "epoch": 0.8837367419091651,
      "grad_norm": 0.31640625,
      "learning_rate": 8.107843045715902e-07,
      "loss": 0.5921,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6499
    },
    {
      "epoch": 0.8838727223279848,
      "grad_norm": 0.361328125,
      "learning_rate": 8.089126325339414e-07,
      "loss": 0.7376,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6500
    },
    {
      "epoch": 0.8840087027468044,
      "grad_norm": 0.314453125,
      "learning_rate": 8.070430322797562e-07,
      "loss": 0.5763,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6501
    },
    {
      "epoch": 0.8841446831656241,
      "grad_norm": 0.447265625,
      "learning_rate": 8.051755042304643e-07,
      "loss": 0.7699,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6502
    },
    {
      "epoch": 0.8842806635844438,
      "grad_norm": 0.41796875,
      "learning_rate": 8.033100488070311e-07,
      "loss": 0.7401,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6503
    },
    {
      "epoch": 0.8844166440032635,
      "grad_norm": 0.5625,
      "learning_rate": 8.014466664299514e-07,
      "loss": 0.6825,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6504
    },
    {
      "epoch": 0.8845526244220833,
      "grad_norm": 0.41015625,
      "learning_rate": 7.995853575192558e-07,
      "loss": 0.8213,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6505
    },
    {
      "epoch": 0.884688604840903,
      "grad_norm": 0.275390625,
      "learning_rate": 7.977261224945055e-07,
      "loss": 0.5088,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6506
    },
    {
      "epoch": 0.8848245852597226,
      "grad_norm": 0.431640625,
      "learning_rate": 7.95868961774795e-07,
      "loss": 0.8187,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6507
    },
    {
      "epoch": 0.8849605656785423,
      "grad_norm": 0.33203125,
      "learning_rate": 7.940138757787507e-07,
      "loss": 0.5951,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6508
    },
    {
      "epoch": 0.885096546097362,
      "grad_norm": 0.2294921875,
      "learning_rate": 7.921608649245327e-07,
      "loss": 0.4312,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6509
    },
    {
      "epoch": 0.8852325265161817,
      "grad_norm": 0.33203125,
      "learning_rate": 7.9030992962983e-07,
      "loss": 0.5851,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6510
    },
    {
      "epoch": 0.8853685069350014,
      "grad_norm": 0.48828125,
      "learning_rate": 7.8846107031187e-07,
      "loss": 0.7643,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6511
    },
    {
      "epoch": 0.885504487353821,
      "grad_norm": 0.2197265625,
      "learning_rate": 7.866142873874028e-07,
      "loss": 0.4206,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6512
    },
    {
      "epoch": 0.8856404677726407,
      "grad_norm": 0.28515625,
      "learning_rate": 7.847695812727218e-07,
      "loss": 0.6047,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6513
    },
    {
      "epoch": 0.8857764481914604,
      "grad_norm": 0.6015625,
      "learning_rate": 7.829269523836413e-07,
      "loss": 0.695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6514
    },
    {
      "epoch": 0.8859124286102801,
      "grad_norm": 0.291015625,
      "learning_rate": 7.810864011355157e-07,
      "loss": 0.5413,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6515
    },
    {
      "epoch": 0.8860484090290998,
      "grad_norm": 0.310546875,
      "learning_rate": 7.792479279432241e-07,
      "loss": 0.6924,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6516
    },
    {
      "epoch": 0.8861843894479196,
      "grad_norm": 0.28515625,
      "learning_rate": 7.774115332211862e-07,
      "loss": 0.5807,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6517
    },
    {
      "epoch": 0.8863203698667392,
      "grad_norm": 0.427734375,
      "learning_rate": 7.75577217383342e-07,
      "loss": 0.5654,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6518
    },
    {
      "epoch": 0.8864563502855589,
      "grad_norm": 0.44921875,
      "learning_rate": 7.737449808431741e-07,
      "loss": 0.7606,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6519
    },
    {
      "epoch": 0.8865923307043786,
      "grad_norm": 0.4296875,
      "learning_rate": 7.719148240136864e-07,
      "loss": 0.7786,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6520
    },
    {
      "epoch": 0.8867283111231983,
      "grad_norm": 0.58984375,
      "learning_rate": 7.700867473074225e-07,
      "loss": 0.6675,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6521
    },
    {
      "epoch": 0.886864291542018,
      "grad_norm": 0.50390625,
      "learning_rate": 7.682607511364492e-07,
      "loss": 0.8175,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6522
    },
    {
      "epoch": 0.8870002719608376,
      "grad_norm": 0.333984375,
      "learning_rate": 7.66436835912373e-07,
      "loss": 0.6632,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6523
    },
    {
      "epoch": 0.8871362523796573,
      "grad_norm": 0.341796875,
      "learning_rate": 7.646150020463228e-07,
      "loss": 0.659,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6524
    },
    {
      "epoch": 0.887272232798477,
      "grad_norm": 0.376953125,
      "learning_rate": 7.627952499489644e-07,
      "loss": 0.5677,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6525
    },
    {
      "epoch": 0.8874082132172967,
      "grad_norm": 0.365234375,
      "learning_rate": 7.609775800304908e-07,
      "loss": 0.5978,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6526
    },
    {
      "epoch": 0.8875441936361164,
      "grad_norm": 0.55078125,
      "learning_rate": 7.591619927006289e-07,
      "loss": 0.5555,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6527
    },
    {
      "epoch": 0.887680174054936,
      "grad_norm": 0.408203125,
      "learning_rate": 7.57348488368631e-07,
      "loss": 0.7122,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6528
    },
    {
      "epoch": 0.8878161544737557,
      "grad_norm": 0.41015625,
      "learning_rate": 7.555370674432872e-07,
      "loss": 0.5527,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6529
    },
    {
      "epoch": 0.8879521348925755,
      "grad_norm": 0.275390625,
      "learning_rate": 7.537277303329105e-07,
      "loss": 0.498,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6530
    },
    {
      "epoch": 0.8880881153113952,
      "grad_norm": 0.578125,
      "learning_rate": 7.519204774453503e-07,
      "loss": 0.5949,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6531
    },
    {
      "epoch": 0.8882240957302149,
      "grad_norm": 0.30078125,
      "learning_rate": 7.501153091879809e-07,
      "loss": 0.5575,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6532
    },
    {
      "epoch": 0.8883600761490346,
      "grad_norm": 0.484375,
      "learning_rate": 7.483122259677111e-07,
      "loss": 0.6444,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6533
    },
    {
      "epoch": 0.8884960565678542,
      "grad_norm": 0.341796875,
      "learning_rate": 7.46511228190977e-07,
      "loss": 0.6642,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6534
    },
    {
      "epoch": 0.8886320369866739,
      "grad_norm": 0.466796875,
      "learning_rate": 7.447123162637482e-07,
      "loss": 0.8478,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6535
    },
    {
      "epoch": 0.8887680174054936,
      "grad_norm": 0.298828125,
      "learning_rate": 7.429154905915159e-07,
      "loss": 0.6243,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6536
    },
    {
      "epoch": 0.8889039978243133,
      "grad_norm": 0.2890625,
      "learning_rate": 7.411207515793128e-07,
      "loss": 0.5081,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6537
    },
    {
      "epoch": 0.889039978243133,
      "grad_norm": 0.2041015625,
      "learning_rate": 7.393280996316899e-07,
      "loss": 0.3722,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6538
    },
    {
      "epoch": 0.8891759586619526,
      "grad_norm": 0.349609375,
      "learning_rate": 7.37537535152737e-07,
      "loss": 0.5845,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6539
    },
    {
      "epoch": 0.8893119390807723,
      "grad_norm": 0.3125,
      "learning_rate": 7.357490585460658e-07,
      "loss": 0.5985,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6540
    },
    {
      "epoch": 0.889447919499592,
      "grad_norm": 0.55078125,
      "learning_rate": 7.339626702148217e-07,
      "loss": 0.5037,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6541
    },
    {
      "epoch": 0.8895838999184118,
      "grad_norm": 0.33203125,
      "learning_rate": 7.321783705616791e-07,
      "loss": 0.6083,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6542
    },
    {
      "epoch": 0.8897198803372315,
      "grad_norm": 0.345703125,
      "learning_rate": 7.303961599888432e-07,
      "loss": 0.5941,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6543
    },
    {
      "epoch": 0.8898558607560512,
      "grad_norm": 0.365234375,
      "learning_rate": 7.286160388980412e-07,
      "loss": 0.6676,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6544
    },
    {
      "epoch": 0.8899918411748708,
      "grad_norm": 0.234375,
      "learning_rate": 7.26838007690539e-07,
      "loss": 0.4528,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6545
    },
    {
      "epoch": 0.8901278215936905,
      "grad_norm": 0.5859375,
      "learning_rate": 7.250620667671227e-07,
      "loss": 0.6081,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6546
    },
    {
      "epoch": 0.8902638020125102,
      "grad_norm": 0.314453125,
      "learning_rate": 7.232882165281141e-07,
      "loss": 0.6265,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6547
    },
    {
      "epoch": 0.8903997824313299,
      "grad_norm": 0.478515625,
      "learning_rate": 7.215164573733569e-07,
      "loss": 0.7483,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6548
    },
    {
      "epoch": 0.8905357628501496,
      "grad_norm": 0.384765625,
      "learning_rate": 7.197467897022325e-07,
      "loss": 0.6336,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6549
    },
    {
      "epoch": 0.8906717432689693,
      "grad_norm": 0.318359375,
      "learning_rate": 7.17979213913641e-07,
      "loss": 0.5975,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6550
    },
    {
      "epoch": 0.8908077236877889,
      "grad_norm": 0.28125,
      "learning_rate": 7.162137304060179e-07,
      "loss": 0.5807,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6551
    },
    {
      "epoch": 0.8909437041066086,
      "grad_norm": 0.5390625,
      "learning_rate": 7.14450339577324e-07,
      "loss": 0.6182,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6552
    },
    {
      "epoch": 0.8910796845254283,
      "grad_norm": 0.451171875,
      "learning_rate": 7.12689041825051e-07,
      "loss": 0.5911,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6553
    },
    {
      "epoch": 0.891215664944248,
      "grad_norm": 0.419921875,
      "learning_rate": 7.109298375462125e-07,
      "loss": 0.6812,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6554
    },
    {
      "epoch": 0.8913516453630678,
      "grad_norm": 0.390625,
      "learning_rate": 7.091727271373606e-07,
      "loss": 0.6847,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6555
    },
    {
      "epoch": 0.8914876257818874,
      "grad_norm": 0.328125,
      "learning_rate": 7.074177109945657e-07,
      "loss": 0.5529,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6556
    },
    {
      "epoch": 0.8916236062007071,
      "grad_norm": 0.404296875,
      "learning_rate": 7.056647895134294e-07,
      "loss": 0.7469,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6557
    },
    {
      "epoch": 0.8917595866195268,
      "grad_norm": 0.56640625,
      "learning_rate": 7.039139630890834e-07,
      "loss": 0.6528,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6558
    },
    {
      "epoch": 0.8918955670383465,
      "grad_norm": 0.392578125,
      "learning_rate": 7.021652321161842e-07,
      "loss": 0.6855,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6559
    },
    {
      "epoch": 0.8920315474571662,
      "grad_norm": 0.75,
      "learning_rate": 7.004185969889188e-07,
      "loss": 0.7334,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6560
    },
    {
      "epoch": 0.8921675278759859,
      "grad_norm": 0.50390625,
      "learning_rate": 6.986740581009977e-07,
      "loss": 0.6462,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6561
    },
    {
      "epoch": 0.8923035082948055,
      "grad_norm": 0.57421875,
      "learning_rate": 6.96931615845664e-07,
      "loss": 0.5313,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6562
    },
    {
      "epoch": 0.8924394887136252,
      "grad_norm": 0.75390625,
      "learning_rate": 6.951912706156838e-07,
      "loss": 0.7546,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6563
    },
    {
      "epoch": 0.8925754691324449,
      "grad_norm": 0.2470703125,
      "learning_rate": 6.934530228033531e-07,
      "loss": 0.4891,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6564
    },
    {
      "epoch": 0.8927114495512646,
      "grad_norm": 0.2734375,
      "learning_rate": 6.91716872800492e-07,
      "loss": 0.5351,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6565
    },
    {
      "epoch": 0.8928474299700843,
      "grad_norm": 0.431640625,
      "learning_rate": 6.899828209984527e-07,
      "loss": 0.7454,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6566
    },
    {
      "epoch": 0.892983410388904,
      "grad_norm": 0.49609375,
      "learning_rate": 6.882508677881094e-07,
      "loss": 0.4898,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6567
    },
    {
      "epoch": 0.8931193908077237,
      "grad_norm": 0.2353515625,
      "learning_rate": 6.865210135598688e-07,
      "loss": 0.5146,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6568
    },
    {
      "epoch": 0.8932553712265434,
      "grad_norm": 0.470703125,
      "learning_rate": 6.847932587036565e-07,
      "loss": 0.6774,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6569
    },
    {
      "epoch": 0.8933913516453631,
      "grad_norm": 0.5234375,
      "learning_rate": 6.830676036089345e-07,
      "loss": 0.6001,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6570
    },
    {
      "epoch": 0.8935273320641828,
      "grad_norm": 0.294921875,
      "learning_rate": 6.813440486646827e-07,
      "loss": 0.5542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6571
    },
    {
      "epoch": 0.8936633124830025,
      "grad_norm": 0.65234375,
      "learning_rate": 6.796225942594148e-07,
      "loss": 0.785,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6572
    },
    {
      "epoch": 0.8937992929018221,
      "grad_norm": 0.62109375,
      "learning_rate": 6.779032407811636e-07,
      "loss": 0.8075,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6573
    },
    {
      "epoch": 0.8939352733206418,
      "grad_norm": 0.380859375,
      "learning_rate": 6.76185988617496e-07,
      "loss": 0.6182,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6574
    },
    {
      "epoch": 0.8940712537394615,
      "grad_norm": 0.349609375,
      "learning_rate": 6.744708381554988e-07,
      "loss": 0.6252,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6575
    },
    {
      "epoch": 0.8942072341582812,
      "grad_norm": 0.453125,
      "learning_rate": 6.727577897817916e-07,
      "loss": 0.748,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6576
    },
    {
      "epoch": 0.8943432145771009,
      "grad_norm": 0.43359375,
      "learning_rate": 6.710468438825124e-07,
      "loss": 0.5687,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6577
    },
    {
      "epoch": 0.8944791949959205,
      "grad_norm": 0.2490234375,
      "learning_rate": 6.693380008433326e-07,
      "loss": 0.5037,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6578
    },
    {
      "epoch": 0.8946151754147402,
      "grad_norm": 0.3828125,
      "learning_rate": 6.676312610494429e-07,
      "loss": 0.3695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6579
    },
    {
      "epoch": 0.89475115583356,
      "grad_norm": 0.1767578125,
      "learning_rate": 6.659266248855667e-07,
      "loss": 0.3066,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6580
    },
    {
      "epoch": 0.8948871362523797,
      "grad_norm": 0.396484375,
      "learning_rate": 6.642240927359478e-07,
      "loss": 0.7882,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6581
    },
    {
      "epoch": 0.8950231166711994,
      "grad_norm": 0.310546875,
      "learning_rate": 6.625236649843592e-07,
      "loss": 0.5846,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6582
    },
    {
      "epoch": 0.8951590970900191,
      "grad_norm": 0.5078125,
      "learning_rate": 6.608253420140953e-07,
      "loss": 0.7298,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6583
    },
    {
      "epoch": 0.8952950775088387,
      "grad_norm": 0.388671875,
      "learning_rate": 6.591291242079822e-07,
      "loss": 0.6989,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6584
    },
    {
      "epoch": 0.8954310579276584,
      "grad_norm": 0.33984375,
      "learning_rate": 6.574350119483664e-07,
      "loss": 0.5871,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6585
    },
    {
      "epoch": 0.8955670383464781,
      "grad_norm": 0.28125,
      "learning_rate": 6.557430056171221e-07,
      "loss": 0.5231,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6586
    },
    {
      "epoch": 0.8957030187652978,
      "grad_norm": 0.2392578125,
      "learning_rate": 6.540531055956478e-07,
      "loss": 0.4191,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6587
    },
    {
      "epoch": 0.8958389991841175,
      "grad_norm": 0.44921875,
      "learning_rate": 6.523653122648687e-07,
      "loss": 0.7065,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6588
    },
    {
      "epoch": 0.8959749796029371,
      "grad_norm": 0.3671875,
      "learning_rate": 6.506796260052306e-07,
      "loss": 0.666,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6589
    },
    {
      "epoch": 0.8961109600217568,
      "grad_norm": 0.298828125,
      "learning_rate": 6.489960471967139e-07,
      "loss": 0.502,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6590
    },
    {
      "epoch": 0.8962469404405765,
      "grad_norm": 0.9453125,
      "learning_rate": 6.473145762188116e-07,
      "loss": 0.6768,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6591
    },
    {
      "epoch": 0.8963829208593962,
      "grad_norm": 0.5625,
      "learning_rate": 6.456352134505517e-07,
      "loss": 0.7152,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6592
    },
    {
      "epoch": 0.896518901278216,
      "grad_norm": 0.326171875,
      "learning_rate": 6.439579592704803e-07,
      "loss": 0.5771,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6593
    },
    {
      "epoch": 0.8966548816970357,
      "grad_norm": 0.32421875,
      "learning_rate": 6.422828140566751e-07,
      "loss": 0.492,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6594
    },
    {
      "epoch": 0.8967908621158553,
      "grad_norm": 0.263671875,
      "learning_rate": 6.406097781867293e-07,
      "loss": 0.5679,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6595
    },
    {
      "epoch": 0.896926842534675,
      "grad_norm": 0.341796875,
      "learning_rate": 6.389388520377704e-07,
      "loss": 0.6556,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6596
    },
    {
      "epoch": 0.8970628229534947,
      "grad_norm": 0.40625,
      "learning_rate": 6.372700359864414e-07,
      "loss": 0.8026,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6597
    },
    {
      "epoch": 0.8971988033723144,
      "grad_norm": 0.486328125,
      "learning_rate": 6.356033304089171e-07,
      "loss": 0.6859,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6598
    },
    {
      "epoch": 0.8973347837911341,
      "grad_norm": 0.376953125,
      "learning_rate": 6.339387356808912e-07,
      "loss": 0.8835,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6599
    },
    {
      "epoch": 0.8974707642099538,
      "grad_norm": 0.6015625,
      "learning_rate": 6.322762521775849e-07,
      "loss": 0.7393,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6600
    },
    {
      "epoch": 0.8976067446287734,
      "grad_norm": 0.37890625,
      "learning_rate": 6.306158802737416e-07,
      "loss": 0.6455,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6601
    },
    {
      "epoch": 0.8977427250475931,
      "grad_norm": 0.6796875,
      "learning_rate": 6.289576203436287e-07,
      "loss": 0.5939,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6602
    },
    {
      "epoch": 0.8978787054664128,
      "grad_norm": 0.451171875,
      "learning_rate": 6.273014727610405e-07,
      "loss": 0.6229,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6603
    },
    {
      "epoch": 0.8980146858852325,
      "grad_norm": 0.5625,
      "learning_rate": 6.256474378992916e-07,
      "loss": 0.5846,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6604
    },
    {
      "epoch": 0.8981506663040523,
      "grad_norm": 0.333984375,
      "learning_rate": 6.239955161312216e-07,
      "loss": 0.548,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6605
    },
    {
      "epoch": 0.898286646722872,
      "grad_norm": 0.4453125,
      "learning_rate": 6.22345707829195e-07,
      "loss": 0.6799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6606
    },
    {
      "epoch": 0.8984226271416916,
      "grad_norm": 0.412109375,
      "learning_rate": 6.206980133650964e-07,
      "loss": 0.6172,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6607
    },
    {
      "epoch": 0.8985586075605113,
      "grad_norm": 0.306640625,
      "learning_rate": 6.190524331103398e-07,
      "loss": 0.5488,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6608
    },
    {
      "epoch": 0.898694587979331,
      "grad_norm": 0.232421875,
      "learning_rate": 6.174089674358563e-07,
      "loss": 0.3825,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6609
    },
    {
      "epoch": 0.8988305683981507,
      "grad_norm": 0.33984375,
      "learning_rate": 6.157676167121018e-07,
      "loss": 0.6644,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6610
    },
    {
      "epoch": 0.8989665488169704,
      "grad_norm": 0.28125,
      "learning_rate": 6.141283813090615e-07,
      "loss": 0.5094,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6611
    },
    {
      "epoch": 0.89910252923579,
      "grad_norm": 0.369140625,
      "learning_rate": 6.124912615962341e-07,
      "loss": 0.6281,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6612
    },
    {
      "epoch": 0.8992385096546097,
      "grad_norm": 0.453125,
      "learning_rate": 6.108562579426502e-07,
      "loss": 0.7875,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6613
    },
    {
      "epoch": 0.8993744900734294,
      "grad_norm": 0.640625,
      "learning_rate": 6.092233707168571e-07,
      "loss": 0.6992,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6614
    },
    {
      "epoch": 0.8995104704922491,
      "grad_norm": 0.447265625,
      "learning_rate": 6.075926002869304e-07,
      "loss": 0.6797,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6615
    },
    {
      "epoch": 0.8996464509110688,
      "grad_norm": 0.380859375,
      "learning_rate": 6.059639470204626e-07,
      "loss": 0.7591,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6616
    },
    {
      "epoch": 0.8997824313298884,
      "grad_norm": 0.359375,
      "learning_rate": 6.043374112845735e-07,
      "loss": 0.6312,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6617
    },
    {
      "epoch": 0.8999184117487082,
      "grad_norm": 0.37109375,
      "learning_rate": 6.027129934459042e-07,
      "loss": 0.6554,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6618
    },
    {
      "epoch": 0.9000543921675279,
      "grad_norm": 0.369140625,
      "learning_rate": 6.010906938706184e-07,
      "loss": 0.6867,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6619
    },
    {
      "epoch": 0.9001903725863476,
      "grad_norm": 0.4296875,
      "learning_rate": 5.994705129244016e-07,
      "loss": 0.5752,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6620
    },
    {
      "epoch": 0.9003263530051673,
      "grad_norm": 0.42578125,
      "learning_rate": 5.978524509724637e-07,
      "loss": 0.4784,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6621
    },
    {
      "epoch": 0.900462333423987,
      "grad_norm": 0.32421875,
      "learning_rate": 5.96236508379533e-07,
      "loss": 0.5459,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6622
    },
    {
      "epoch": 0.9005983138428066,
      "grad_norm": 0.55859375,
      "learning_rate": 5.946226855098658e-07,
      "loss": 0.6551,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6623
    },
    {
      "epoch": 0.9007342942616263,
      "grad_norm": 0.3984375,
      "learning_rate": 5.930109827272357e-07,
      "loss": 0.5892,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6624
    },
    {
      "epoch": 0.900870274680446,
      "grad_norm": 0.42578125,
      "learning_rate": 5.914014003949408e-07,
      "loss": 0.8079,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6625
    },
    {
      "epoch": 0.9010062550992657,
      "grad_norm": 0.447265625,
      "learning_rate": 5.897939388758001e-07,
      "loss": 0.7288,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6626
    },
    {
      "epoch": 0.9011422355180854,
      "grad_norm": 0.3828125,
      "learning_rate": 5.881885985321567e-07,
      "loss": 0.8478,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6627
    },
    {
      "epoch": 0.901278215936905,
      "grad_norm": 0.546875,
      "learning_rate": 5.865853797258714e-07,
      "loss": 0.5524,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6628
    },
    {
      "epoch": 0.9014141963557247,
      "grad_norm": 0.2890625,
      "learning_rate": 5.849842828183316e-07,
      "loss": 0.5814,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6629
    },
    {
      "epoch": 0.9015501767745445,
      "grad_norm": 0.26953125,
      "learning_rate": 5.833853081704422e-07,
      "loss": 0.4783,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6630
    },
    {
      "epoch": 0.9016861571933642,
      "grad_norm": 0.55078125,
      "learning_rate": 5.817884561426346e-07,
      "loss": 0.5203,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6631
    },
    {
      "epoch": 0.9018221376121839,
      "grad_norm": 0.306640625,
      "learning_rate": 5.801937270948543e-07,
      "loss": 0.5448,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6632
    },
    {
      "epoch": 0.9019581180310036,
      "grad_norm": 0.6953125,
      "learning_rate": 5.786011213865772e-07,
      "loss": 0.73,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6633
    },
    {
      "epoch": 0.9020940984498232,
      "grad_norm": 0.59375,
      "learning_rate": 5.770106393767926e-07,
      "loss": 0.7913,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6634
    },
    {
      "epoch": 0.9022300788686429,
      "grad_norm": 0.515625,
      "learning_rate": 5.754222814240174e-07,
      "loss": 0.7686,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6635
    },
    {
      "epoch": 0.9023660592874626,
      "grad_norm": 0.33984375,
      "learning_rate": 5.738360478862848e-07,
      "loss": 0.5285,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6636
    },
    {
      "epoch": 0.9025020397062823,
      "grad_norm": 0.3203125,
      "learning_rate": 5.722519391211523e-07,
      "loss": 0.5588,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6637
    },
    {
      "epoch": 0.902638020125102,
      "grad_norm": 0.275390625,
      "learning_rate": 5.706699554856965e-07,
      "loss": 0.4796,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6638
    },
    {
      "epoch": 0.9027740005439217,
      "grad_norm": 0.44921875,
      "learning_rate": 5.690900973365177e-07,
      "loss": 0.7401,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6639
    },
    {
      "epoch": 0.9029099809627413,
      "grad_norm": 0.408203125,
      "learning_rate": 5.675123650297332e-07,
      "loss": 0.7985,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6640
    },
    {
      "epoch": 0.903045961381561,
      "grad_norm": 0.6171875,
      "learning_rate": 5.659367589209841e-07,
      "loss": 0.6401,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6641
    },
    {
      "epoch": 0.9031819418003807,
      "grad_norm": 0.33203125,
      "learning_rate": 5.643632793654308e-07,
      "loss": 0.5269,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6642
    },
    {
      "epoch": 0.9033179222192005,
      "grad_norm": 0.50390625,
      "learning_rate": 5.627919267177562e-07,
      "loss": 0.741,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6643
    },
    {
      "epoch": 0.9034539026380202,
      "grad_norm": 0.384765625,
      "learning_rate": 5.612227013321603e-07,
      "loss": 0.8535,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6644
    },
    {
      "epoch": 0.9035898830568398,
      "grad_norm": 0.609375,
      "learning_rate": 5.596556035623679e-07,
      "loss": 0.715,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6645
    },
    {
      "epoch": 0.9037258634756595,
      "grad_norm": 0.5078125,
      "learning_rate": 5.580906337616198e-07,
      "loss": 0.6323,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6646
    },
    {
      "epoch": 0.9038618438944792,
      "grad_norm": 0.32421875,
      "learning_rate": 5.565277922826806e-07,
      "loss": 0.6506,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6647
    },
    {
      "epoch": 0.9039978243132989,
      "grad_norm": 0.498046875,
      "learning_rate": 5.549670794778339e-07,
      "loss": 0.6198,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6648
    },
    {
      "epoch": 0.9041338047321186,
      "grad_norm": 0.55078125,
      "learning_rate": 5.534084956988839e-07,
      "loss": 0.6496,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6649
    },
    {
      "epoch": 0.9042697851509383,
      "grad_norm": 0.51171875,
      "learning_rate": 5.518520412971517e-07,
      "loss": 0.6693,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6650
    },
    {
      "epoch": 0.9044057655697579,
      "grad_norm": 0.2734375,
      "learning_rate": 5.502977166234857e-07,
      "loss": 0.5431,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6651
    },
    {
      "epoch": 0.9045417459885776,
      "grad_norm": 0.5234375,
      "learning_rate": 5.487455220282445e-07,
      "loss": 0.7413,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6652
    },
    {
      "epoch": 0.9046777264073973,
      "grad_norm": 0.37890625,
      "learning_rate": 5.47195457861317e-07,
      "loss": 0.7171,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6653
    },
    {
      "epoch": 0.904813706826217,
      "grad_norm": 0.453125,
      "learning_rate": 5.456475244721016e-07,
      "loss": 0.8162,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6654
    },
    {
      "epoch": 0.9049496872450367,
      "grad_norm": 0.31640625,
      "learning_rate": 5.441017222095268e-07,
      "loss": 0.6029,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6655
    },
    {
      "epoch": 0.9050856676638565,
      "grad_norm": 0.36328125,
      "learning_rate": 5.425580514220297e-07,
      "loss": 0.6496,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6656
    },
    {
      "epoch": 0.9052216480826761,
      "grad_norm": 0.77734375,
      "learning_rate": 5.410165124575772e-07,
      "loss": 0.7017,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6657
    },
    {
      "epoch": 0.9053576285014958,
      "grad_norm": 0.31640625,
      "learning_rate": 5.394771056636472e-07,
      "loss": 0.5247,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6658
    },
    {
      "epoch": 0.9054936089203155,
      "grad_norm": 0.435546875,
      "learning_rate": 5.379398313872441e-07,
      "loss": 0.8184,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6659
    },
    {
      "epoch": 0.9056295893391352,
      "grad_norm": 0.73828125,
      "learning_rate": 5.364046899748865e-07,
      "loss": 0.7783,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6660
    },
    {
      "epoch": 0.9057655697579549,
      "grad_norm": 0.376953125,
      "learning_rate": 5.34871681772614e-07,
      "loss": 0.7803,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6661
    },
    {
      "epoch": 0.9059015501767745,
      "grad_norm": 0.6796875,
      "learning_rate": 5.333408071259849e-07,
      "loss": 0.8407,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6662
    },
    {
      "epoch": 0.9060375305955942,
      "grad_norm": 0.400390625,
      "learning_rate": 5.318120663800796e-07,
      "loss": 0.5251,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6663
    },
    {
      "epoch": 0.9061735110144139,
      "grad_norm": 0.263671875,
      "learning_rate": 5.302854598794938e-07,
      "loss": 0.4285,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6664
    },
    {
      "epoch": 0.9063094914332336,
      "grad_norm": 0.380859375,
      "learning_rate": 5.28760987968342e-07,
      "loss": 0.6833,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6665
    },
    {
      "epoch": 0.9064454718520533,
      "grad_norm": 0.310546875,
      "learning_rate": 5.272386509902605e-07,
      "loss": 0.5934,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6666
    },
    {
      "epoch": 0.906581452270873,
      "grad_norm": 0.431640625,
      "learning_rate": 5.257184492884004e-07,
      "loss": 0.8154,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6667
    },
    {
      "epoch": 0.9067174326896927,
      "grad_norm": 0.482421875,
      "learning_rate": 5.242003832054365e-07,
      "loss": 0.7008,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6668
    },
    {
      "epoch": 0.9068534131085124,
      "grad_norm": 0.53515625,
      "learning_rate": 5.226844530835562e-07,
      "loss": 0.5366,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6669
    },
    {
      "epoch": 0.9069893935273321,
      "grad_norm": 0.2275390625,
      "learning_rate": 5.211706592644728e-07,
      "loss": 0.4471,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6670
    },
    {
      "epoch": 0.9071253739461518,
      "grad_norm": 0.66015625,
      "learning_rate": 5.196590020894098e-07,
      "loss": 0.7007,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6671
    },
    {
      "epoch": 0.9072613543649715,
      "grad_norm": 0.2890625,
      "learning_rate": 5.181494818991173e-07,
      "loss": 0.5208,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6672
    },
    {
      "epoch": 0.9073973347837911,
      "grad_norm": 0.578125,
      "learning_rate": 5.166420990338549e-07,
      "loss": 0.8296,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6673
    },
    {
      "epoch": 0.9075333152026108,
      "grad_norm": 0.359375,
      "learning_rate": 5.1513685383341e-07,
      "loss": 0.5254,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6674
    },
    {
      "epoch": 0.9076692956214305,
      "grad_norm": 0.46875,
      "learning_rate": 5.136337466370788e-07,
      "loss": 0.5959,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6675
    },
    {
      "epoch": 0.9078052760402502,
      "grad_norm": 0.2734375,
      "learning_rate": 5.121327777836849e-07,
      "loss": 0.5033,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6676
    },
    {
      "epoch": 0.9079412564590699,
      "grad_norm": 0.28515625,
      "learning_rate": 5.106339476115596e-07,
      "loss": 0.5993,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6677
    },
    {
      "epoch": 0.9080772368778895,
      "grad_norm": 0.359375,
      "learning_rate": 5.091372564585617e-07,
      "loss": 0.6851,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6678
    },
    {
      "epoch": 0.9082132172967092,
      "grad_norm": 0.66796875,
      "learning_rate": 5.076427046620613e-07,
      "loss": 0.5576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6679
    },
    {
      "epoch": 0.9083491977155289,
      "grad_norm": 0.3984375,
      "learning_rate": 5.061502925589512e-07,
      "loss": 0.6714,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6680
    },
    {
      "epoch": 0.9084851781343487,
      "grad_norm": 0.263671875,
      "learning_rate": 5.046600204856356e-07,
      "loss": 0.5042,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6681
    },
    {
      "epoch": 0.9086211585531684,
      "grad_norm": 0.283203125,
      "learning_rate": 5.031718887780445e-07,
      "loss": 0.4622,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6682
    },
    {
      "epoch": 0.9087571389719881,
      "grad_norm": 0.3359375,
      "learning_rate": 5.016858977716166e-07,
      "loss": 0.5394,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6683
    },
    {
      "epoch": 0.9088931193908077,
      "grad_norm": 0.40625,
      "learning_rate": 5.002020478013158e-07,
      "loss": 0.7362,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6684
    },
    {
      "epoch": 0.9090290998096274,
      "grad_norm": 0.48046875,
      "learning_rate": 4.987203392016183e-07,
      "loss": 0.7021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6685
    },
    {
      "epoch": 0.9091650802284471,
      "grad_norm": 0.51171875,
      "learning_rate": 4.972407723065198e-07,
      "loss": 0.4089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6686
    },
    {
      "epoch": 0.9093010606472668,
      "grad_norm": 0.6796875,
      "learning_rate": 4.957633474495316e-07,
      "loss": 0.861,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6687
    },
    {
      "epoch": 0.9094370410660865,
      "grad_norm": 0.34375,
      "learning_rate": 4.942880649636861e-07,
      "loss": 0.6047,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6688
    },
    {
      "epoch": 0.9095730214849062,
      "grad_norm": 0.73828125,
      "learning_rate": 4.928149251815273e-07,
      "loss": 0.6532,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6689
    },
    {
      "epoch": 0.9097090019037258,
      "grad_norm": 0.423828125,
      "learning_rate": 4.913439284351207e-07,
      "loss": 0.6948,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6690
    },
    {
      "epoch": 0.9098449823225455,
      "grad_norm": 0.31640625,
      "learning_rate": 4.898750750560455e-07,
      "loss": 0.6917,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6691
    },
    {
      "epoch": 0.9099809627413652,
      "grad_norm": 0.384765625,
      "learning_rate": 4.884083653754013e-07,
      "loss": 0.673,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6692
    },
    {
      "epoch": 0.910116943160185,
      "grad_norm": 0.64453125,
      "learning_rate": 4.869437997237991e-07,
      "loss": 0.6831,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6693
    },
    {
      "epoch": 0.9102529235790047,
      "grad_norm": 0.224609375,
      "learning_rate": 4.854813784313728e-07,
      "loss": 0.4476,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6694
    },
    {
      "epoch": 0.9103889039978244,
      "grad_norm": 0.494140625,
      "learning_rate": 4.840211018277685e-07,
      "loss": 0.877,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6695
    },
    {
      "epoch": 0.910524884416644,
      "grad_norm": 0.3671875,
      "learning_rate": 4.82562970242152e-07,
      "loss": 0.6673,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6696
    },
    {
      "epoch": 0.9106608648354637,
      "grad_norm": 0.39453125,
      "learning_rate": 4.811069840032023e-07,
      "loss": 0.7585,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6697
    },
    {
      "epoch": 0.9107968452542834,
      "grad_norm": 0.279296875,
      "learning_rate": 4.796531434391171e-07,
      "loss": 0.4884,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6698
    },
    {
      "epoch": 0.9109328256731031,
      "grad_norm": 0.2890625,
      "learning_rate": 4.782014488776099e-07,
      "loss": 0.5296,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6699
    },
    {
      "epoch": 0.9110688060919228,
      "grad_norm": 0.2490234375,
      "learning_rate": 4.7675190064591094e-07,
      "loss": 0.4813,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6700
    },
    {
      "epoch": 0.9112047865107424,
      "grad_norm": 0.3671875,
      "learning_rate": 4.7530449907076446e-07,
      "loss": 0.7147,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6701
    },
    {
      "epoch": 0.9113407669295621,
      "grad_norm": 0.41796875,
      "learning_rate": 4.738592444784351e-07,
      "loss": 0.6709,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6702
    },
    {
      "epoch": 0.9114767473483818,
      "grad_norm": 0.39453125,
      "learning_rate": 4.7241613719469784e-07,
      "loss": 0.7194,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6703
    },
    {
      "epoch": 0.9116127277672015,
      "grad_norm": 0.3984375,
      "learning_rate": 4.709751775448501e-07,
      "loss": 0.5939,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6704
    },
    {
      "epoch": 0.9117487081860212,
      "grad_norm": 0.359375,
      "learning_rate": 4.695363658536978e-07,
      "loss": 0.6548,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6705
    },
    {
      "epoch": 0.911884688604841,
      "grad_norm": 0.44921875,
      "learning_rate": 4.6809970244557024e-07,
      "loss": 0.7629,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6706
    },
    {
      "epoch": 0.9120206690236606,
      "grad_norm": 0.40234375,
      "learning_rate": 4.6666518764430736e-07,
      "loss": 0.7643,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6707
    },
    {
      "epoch": 0.9121566494424803,
      "grad_norm": 0.32421875,
      "learning_rate": 4.6523282177326713e-07,
      "loss": 0.6688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6708
    },
    {
      "epoch": 0.9122926298613,
      "grad_norm": 0.404296875,
      "learning_rate": 4.6380260515532015e-07,
      "loss": 0.7946,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6709
    },
    {
      "epoch": 0.9124286102801197,
      "grad_norm": 0.43359375,
      "learning_rate": 4.623745381128575e-07,
      "loss": 0.6359,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6710
    },
    {
      "epoch": 0.9125645906989394,
      "grad_norm": 0.6171875,
      "learning_rate": 4.609486209677805e-07,
      "loss": 0.7899,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6711
    },
    {
      "epoch": 0.912700571117759,
      "grad_norm": 0.51171875,
      "learning_rate": 4.595248540415109e-07,
      "loss": 0.6543,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6712
    },
    {
      "epoch": 0.9128365515365787,
      "grad_norm": 0.373046875,
      "learning_rate": 4.5810323765497987e-07,
      "loss": 0.6484,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6713
    },
    {
      "epoch": 0.9129725319553984,
      "grad_norm": 0.3984375,
      "learning_rate": 4.56683772128641e-07,
      "loss": 0.7397,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6714
    },
    {
      "epoch": 0.9131085123742181,
      "grad_norm": 0.462890625,
      "learning_rate": 4.5526645778245615e-07,
      "loss": 0.5799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6715
    },
    {
      "epoch": 0.9132444927930378,
      "grad_norm": 0.3671875,
      "learning_rate": 4.538512949359075e-07,
      "loss": 0.5581,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6716
    },
    {
      "epoch": 0.9133804732118574,
      "grad_norm": 0.384765625,
      "learning_rate": 4.5243828390798773e-07,
      "loss": 0.627,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6717
    },
    {
      "epoch": 0.9135164536306771,
      "grad_norm": 0.78515625,
      "learning_rate": 4.5102742501720976e-07,
      "loss": 0.8825,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6718
    },
    {
      "epoch": 0.9136524340494969,
      "grad_norm": 0.60546875,
      "learning_rate": 4.4961871858159477e-07,
      "loss": 0.4889,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6719
    },
    {
      "epoch": 0.9137884144683166,
      "grad_norm": 0.80859375,
      "learning_rate": 4.4821216491868545e-07,
      "loss": 0.6566,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6720
    },
    {
      "epoch": 0.9139243948871363,
      "grad_norm": 0.322265625,
      "learning_rate": 4.4680776434553375e-07,
      "loss": 0.6045,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6721
    },
    {
      "epoch": 0.914060375305956,
      "grad_norm": 0.51953125,
      "learning_rate": 4.4540551717871083e-07,
      "loss": 0.7588,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6722
    },
    {
      "epoch": 0.9141963557247756,
      "grad_norm": 0.30078125,
      "learning_rate": 4.4400542373429946e-07,
      "loss": 0.4552,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6723
    },
    {
      "epoch": 0.9143323361435953,
      "grad_norm": 0.265625,
      "learning_rate": 4.4260748432789826e-07,
      "loss": 0.4927,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6724
    },
    {
      "epoch": 0.914468316562415,
      "grad_norm": 0.474609375,
      "learning_rate": 4.412116992746196e-07,
      "loss": 0.6771,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6725
    },
    {
      "epoch": 0.9146042969812347,
      "grad_norm": 0.287109375,
      "learning_rate": 4.398180688890896e-07,
      "loss": 0.4904,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6726
    },
    {
      "epoch": 0.9147402774000544,
      "grad_norm": 0.6875,
      "learning_rate": 4.384265934854526e-07,
      "loss": 0.6587,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6727
    },
    {
      "epoch": 0.914876257818874,
      "grad_norm": 0.828125,
      "learning_rate": 4.370372733773598e-07,
      "loss": 0.8351,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6728
    },
    {
      "epoch": 0.9150122382376937,
      "grad_norm": 0.2890625,
      "learning_rate": 4.356501088779841e-07,
      "loss": 0.4793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6729
    },
    {
      "epoch": 0.9151482186565134,
      "grad_norm": 0.271484375,
      "learning_rate": 4.342651003000076e-07,
      "loss": 0.5465,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6730
    },
    {
      "epoch": 0.9152841990753332,
      "grad_norm": 0.55859375,
      "learning_rate": 4.328822479556294e-07,
      "loss": 0.4238,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6731
    },
    {
      "epoch": 0.9154201794941529,
      "grad_norm": 0.291015625,
      "learning_rate": 4.315015521565602e-07,
      "loss": 0.6592,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6732
    },
    {
      "epoch": 0.9155561599129726,
      "grad_norm": 0.375,
      "learning_rate": 4.301230132140277e-07,
      "loss": 0.5977,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6733
    },
    {
      "epoch": 0.9156921403317922,
      "grad_norm": 1.078125,
      "learning_rate": 4.287466314387689e-07,
      "loss": 0.8211,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6734
    },
    {
      "epoch": 0.9158281207506119,
      "grad_norm": 0.220703125,
      "learning_rate": 4.2737240714104014e-07,
      "loss": 0.3625,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6735
    },
    {
      "epoch": 0.9159641011694316,
      "grad_norm": 0.63671875,
      "learning_rate": 4.260003406306046e-07,
      "loss": 0.5654,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6736
    },
    {
      "epoch": 0.9161000815882513,
      "grad_norm": 0.330078125,
      "learning_rate": 4.2463043221674603e-07,
      "loss": 0.6004,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6737
    },
    {
      "epoch": 0.916236062007071,
      "grad_norm": 0.64453125,
      "learning_rate": 4.2326268220825526e-07,
      "loss": 0.6478,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6738
    },
    {
      "epoch": 0.9163720424258907,
      "grad_norm": 0.291015625,
      "learning_rate": 4.2189709091344453e-07,
      "loss": 0.5075,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6739
    },
    {
      "epoch": 0.9165080228447103,
      "grad_norm": 0.26953125,
      "learning_rate": 4.2053365864012985e-07,
      "loss": 0.4961,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6740
    },
    {
      "epoch": 0.91664400326353,
      "grad_norm": 0.44921875,
      "learning_rate": 4.1917238569564977e-07,
      "loss": 0.5884,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6741
    },
    {
      "epoch": 0.9167799836823497,
      "grad_norm": 0.244140625,
      "learning_rate": 4.1781327238684775e-07,
      "loss": 0.5358,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6742
    },
    {
      "epoch": 0.9169159641011694,
      "grad_norm": 0.5234375,
      "learning_rate": 4.1645631902008876e-07,
      "loss": 0.4272,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6743
    },
    {
      "epoch": 0.9170519445199892,
      "grad_norm": 0.404296875,
      "learning_rate": 4.151015259012436e-07,
      "loss": 0.798,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6744
    },
    {
      "epoch": 0.9171879249388089,
      "grad_norm": 0.369140625,
      "learning_rate": 4.1374889333570035e-07,
      "loss": 0.6893,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6745
    },
    {
      "epoch": 0.9173239053576285,
      "grad_norm": 0.5234375,
      "learning_rate": 4.123984216283583e-07,
      "loss": 0.5495,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6746
    },
    {
      "epoch": 0.9174598857764482,
      "grad_norm": 0.455078125,
      "learning_rate": 4.110501110836329e-07,
      "loss": 0.5024,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6747
    },
    {
      "epoch": 0.9175958661952679,
      "grad_norm": 0.263671875,
      "learning_rate": 4.0970396200544546e-07,
      "loss": 0.4842,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6748
    },
    {
      "epoch": 0.9177318466140876,
      "grad_norm": 0.65625,
      "learning_rate": 4.0835997469723776e-07,
      "loss": 0.585,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6749
    },
    {
      "epoch": 0.9178678270329073,
      "grad_norm": 0.3515625,
      "learning_rate": 4.070181494619596e-07,
      "loss": 0.6805,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6750
    },
    {
      "epoch": 0.9180038074517269,
      "grad_norm": 0.400390625,
      "learning_rate": 4.0567848660207576e-07,
      "loss": 0.8441,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6751
    },
    {
      "epoch": 0.9181397878705466,
      "grad_norm": 0.59765625,
      "learning_rate": 4.043409864195602e-07,
      "loss": 0.7503,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6752
    },
    {
      "epoch": 0.9182757682893663,
      "grad_norm": 0.59765625,
      "learning_rate": 4.0300564921590515e-07,
      "loss": 0.4741,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6753
    },
    {
      "epoch": 0.918411748708186,
      "grad_norm": 0.3671875,
      "learning_rate": 4.0167247529210974e-07,
      "loss": 0.5643,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6754
    },
    {
      "epoch": 0.9185477291270057,
      "grad_norm": 0.37109375,
      "learning_rate": 4.003414649486892e-07,
      "loss": 0.6657,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6755
    },
    {
      "epoch": 0.9186837095458255,
      "grad_norm": 0.56640625,
      "learning_rate": 3.9901261848566687e-07,
      "loss": 0.8073,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6756
    },
    {
      "epoch": 0.9188196899646451,
      "grad_norm": 0.302734375,
      "learning_rate": 3.9768593620258533e-07,
      "loss": 0.596,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6757
    },
    {
      "epoch": 0.9189556703834648,
      "grad_norm": 0.30859375,
      "learning_rate": 3.9636141839848987e-07,
      "loss": 0.5449,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6758
    },
    {
      "epoch": 0.9190916508022845,
      "grad_norm": 0.3125,
      "learning_rate": 3.950390653719471e-07,
      "loss": 0.5545,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6759
    },
    {
      "epoch": 0.9192276312211042,
      "grad_norm": 0.421875,
      "learning_rate": 3.9371887742102986e-07,
      "loss": 0.6292,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6760
    },
    {
      "epoch": 0.9193636116399239,
      "grad_norm": 0.345703125,
      "learning_rate": 3.9240085484332445e-07,
      "loss": 0.6925,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6761
    },
    {
      "epoch": 0.9194995920587435,
      "grad_norm": 0.458984375,
      "learning_rate": 3.910849979359288e-07,
      "loss": 0.6439,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6762
    },
    {
      "epoch": 0.9196355724775632,
      "grad_norm": 0.421875,
      "learning_rate": 3.897713069954556e-07,
      "loss": 0.6143,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6763
    },
    {
      "epoch": 0.9197715528963829,
      "grad_norm": 0.341796875,
      "learning_rate": 3.8845978231802474e-07,
      "loss": 0.7651,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6764
    },
    {
      "epoch": 0.9199075333152026,
      "grad_norm": 0.400390625,
      "learning_rate": 3.871504241992707e-07,
      "loss": 0.5392,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6765
    },
    {
      "epoch": 0.9200435137340223,
      "grad_norm": 0.359375,
      "learning_rate": 3.858432329343376e-07,
      "loss": 0.6174,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6766
    },
    {
      "epoch": 0.920179494152842,
      "grad_norm": 0.71875,
      "learning_rate": 3.8453820881788394e-07,
      "loss": 0.9759,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6767
    },
    {
      "epoch": 0.9203154745716616,
      "grad_norm": 0.26171875,
      "learning_rate": 3.832353521440768e-07,
      "loss": 0.3973,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6768
    },
    {
      "epoch": 0.9204514549904814,
      "grad_norm": 0.439453125,
      "learning_rate": 3.819346632065968e-07,
      "loss": 0.6771,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6769
    },
    {
      "epoch": 0.9205874354093011,
      "grad_norm": 0.80859375,
      "learning_rate": 3.806361422986349e-07,
      "loss": 0.7038,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6770
    },
    {
      "epoch": 0.9207234158281208,
      "grad_norm": 0.58203125,
      "learning_rate": 3.793397897128948e-07,
      "loss": 0.6271,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6771
    },
    {
      "epoch": 0.9208593962469405,
      "grad_norm": 0.380859375,
      "learning_rate": 3.780456057415871e-07,
      "loss": 0.5913,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6772
    },
    {
      "epoch": 0.9209953766657601,
      "grad_norm": 0.376953125,
      "learning_rate": 3.767535906764397e-07,
      "loss": 0.7765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6773
    },
    {
      "epoch": 0.9211313570845798,
      "grad_norm": 0.4921875,
      "learning_rate": 3.754637448086873e-07,
      "loss": 0.7321,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6774
    },
    {
      "epoch": 0.9212673375033995,
      "grad_norm": 0.3828125,
      "learning_rate": 3.7417606842907847e-07,
      "loss": 0.5703,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6775
    },
    {
      "epoch": 0.9214033179222192,
      "grad_norm": 0.30078125,
      "learning_rate": 3.7289056182786774e-07,
      "loss": 0.5381,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6776
    },
    {
      "epoch": 0.9215392983410389,
      "grad_norm": 0.455078125,
      "learning_rate": 3.7160722529482773e-07,
      "loss": 0.4682,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6777
    },
    {
      "epoch": 0.9216752787598586,
      "grad_norm": 0.435546875,
      "learning_rate": 3.7032605911923593e-07,
      "loss": 0.5815,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6778
    },
    {
      "epoch": 0.9218112591786782,
      "grad_norm": 0.388671875,
      "learning_rate": 3.690470635898835e-07,
      "loss": 0.6649,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6779
    },
    {
      "epoch": 0.9219472395974979,
      "grad_norm": 0.462890625,
      "learning_rate": 3.6777023899507216e-07,
      "loss": 0.8081,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6780
    },
    {
      "epoch": 0.9220832200163176,
      "grad_norm": 0.73828125,
      "learning_rate": 3.6649558562261377e-07,
      "loss": 0.6405,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6781
    },
    {
      "epoch": 0.9222192004351374,
      "grad_norm": 0.67578125,
      "learning_rate": 3.652231037598286e-07,
      "loss": 0.8717,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6782
    },
    {
      "epoch": 0.9223551808539571,
      "grad_norm": 0.314453125,
      "learning_rate": 3.6395279369355274e-07,
      "loss": 0.7306,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6783
    },
    {
      "epoch": 0.9224911612727767,
      "grad_norm": 0.45703125,
      "learning_rate": 3.6268465571012824e-07,
      "loss": 0.4967,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6784
    },
    {
      "epoch": 0.9226271416915964,
      "grad_norm": 0.388671875,
      "learning_rate": 3.6141869009540866e-07,
      "loss": 0.4352,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6785
    },
    {
      "epoch": 0.9227631221104161,
      "grad_norm": 0.408203125,
      "learning_rate": 3.6015489713475906e-07,
      "loss": 0.5957,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6786
    },
    {
      "epoch": 0.9228991025292358,
      "grad_norm": 0.291015625,
      "learning_rate": 3.5889327711305156e-07,
      "loss": 0.515,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6787
    },
    {
      "epoch": 0.9230350829480555,
      "grad_norm": 0.310546875,
      "learning_rate": 3.5763383031467425e-07,
      "loss": 0.5345,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6788
    },
    {
      "epoch": 0.9231710633668752,
      "grad_norm": 0.431640625,
      "learning_rate": 3.563765570235178e-07,
      "loss": 0.8464,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6789
    },
    {
      "epoch": 0.9233070437856948,
      "grad_norm": 1.125,
      "learning_rate": 3.551214575229911e-07,
      "loss": 0.9591,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6790
    },
    {
      "epoch": 0.9234430242045145,
      "grad_norm": 0.365234375,
      "learning_rate": 3.5386853209600557e-07,
      "loss": 0.5098,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6791
    },
    {
      "epoch": 0.9235790046233342,
      "grad_norm": 0.3359375,
      "learning_rate": 3.526177810249887e-07,
      "loss": 0.646,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6792
    },
    {
      "epoch": 0.9237149850421539,
      "grad_norm": 0.302734375,
      "learning_rate": 3.513692045918726e-07,
      "loss": 0.6146,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6793
    },
    {
      "epoch": 0.9238509654609737,
      "grad_norm": 0.5703125,
      "learning_rate": 3.5012280307810344e-07,
      "loss": 0.739,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6794
    },
    {
      "epoch": 0.9239869458797934,
      "grad_norm": 0.39453125,
      "learning_rate": 3.488785767646341e-07,
      "loss": 0.6955,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6795
    },
    {
      "epoch": 0.924122926298613,
      "grad_norm": 0.8984375,
      "learning_rate": 3.476365259319292e-07,
      "loss": 0.6956,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6796
    },
    {
      "epoch": 0.9242589067174327,
      "grad_norm": 0.3046875,
      "learning_rate": 3.463966508599614e-07,
      "loss": 0.6087,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6797
    },
    {
      "epoch": 0.9243948871362524,
      "grad_norm": 0.3203125,
      "learning_rate": 3.45158951828215e-07,
      "loss": 0.7153,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6798
    },
    {
      "epoch": 0.9245308675550721,
      "grad_norm": 0.59765625,
      "learning_rate": 3.439234291156801e-07,
      "loss": 0.4855,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6799
    },
    {
      "epoch": 0.9246668479738918,
      "grad_norm": 0.53515625,
      "learning_rate": 3.426900830008628e-07,
      "loss": 0.7345,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6800
    },
    {
      "epoch": 0.9248028283927114,
      "grad_norm": 0.353515625,
      "learning_rate": 3.414589137617697e-07,
      "loss": 0.7212,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6801
    },
    {
      "epoch": 0.9249388088115311,
      "grad_norm": 0.482421875,
      "learning_rate": 3.402299216759242e-07,
      "loss": 0.5355,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6802
    },
    {
      "epoch": 0.9250747892303508,
      "grad_norm": 0.373046875,
      "learning_rate": 3.390031070203548e-07,
      "loss": 0.2656,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6803
    },
    {
      "epoch": 0.9252107696491705,
      "grad_norm": 0.28515625,
      "learning_rate": 3.3777847007160245e-07,
      "loss": 0.4932,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6804
    },
    {
      "epoch": 0.9253467500679902,
      "grad_norm": 0.263671875,
      "learning_rate": 3.365560111057131e-07,
      "loss": 0.3849,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6805
    },
    {
      "epoch": 0.9254827304868098,
      "grad_norm": 0.498046875,
      "learning_rate": 3.353357303982463e-07,
      "loss": 0.665,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6806
    },
    {
      "epoch": 0.9256187109056296,
      "grad_norm": 0.5546875,
      "learning_rate": 3.341176282242653e-07,
      "loss": 0.6314,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6807
    },
    {
      "epoch": 0.9257546913244493,
      "grad_norm": 0.345703125,
      "learning_rate": 3.3290170485834715e-07,
      "loss": 0.5746,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6808
    },
    {
      "epoch": 0.925890671743269,
      "grad_norm": 0.357421875,
      "learning_rate": 3.31687960574576e-07,
      "loss": 0.5602,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6809
    },
    {
      "epoch": 0.9260266521620887,
      "grad_norm": 0.40234375,
      "learning_rate": 3.30476395646544e-07,
      "loss": 0.7596,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6810
    },
    {
      "epoch": 0.9261626325809084,
      "grad_norm": 0.251953125,
      "learning_rate": 3.2926701034735163e-07,
      "loss": 0.4993,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6811
    },
    {
      "epoch": 0.926298612999728,
      "grad_norm": 0.5703125,
      "learning_rate": 3.28059804949612e-07,
      "loss": 0.4497,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6812
    },
    {
      "epoch": 0.9264345934185477,
      "grad_norm": 0.2421875,
      "learning_rate": 3.2685477972544064e-07,
      "loss": 0.4777,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6813
    },
    {
      "epoch": 0.9265705738373674,
      "grad_norm": 0.59765625,
      "learning_rate": 3.256519349464671e-07,
      "loss": 0.7295,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6814
    },
    {
      "epoch": 0.9267065542561871,
      "grad_norm": 0.26171875,
      "learning_rate": 3.2445127088382655e-07,
      "loss": 0.5042,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6815
    },
    {
      "epoch": 0.9268425346750068,
      "grad_norm": 0.51953125,
      "learning_rate": 3.232527878081648e-07,
      "loss": 0.5396,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6816
    },
    {
      "epoch": 0.9269785150938264,
      "grad_norm": 0.34375,
      "learning_rate": 3.2205648598963245e-07,
      "loss": 0.6468,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6817
    },
    {
      "epoch": 0.9271144955126461,
      "grad_norm": 3.8125,
      "learning_rate": 3.2086236569789155e-07,
      "loss": 0.7173,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6818
    },
    {
      "epoch": 0.9272504759314659,
      "grad_norm": 0.294921875,
      "learning_rate": 3.196704272021112e-07,
      "loss": 0.418,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6819
    },
    {
      "epoch": 0.9273864563502856,
      "grad_norm": 0.271484375,
      "learning_rate": 3.184806707709698e-07,
      "loss": 0.4998,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6820
    },
    {
      "epoch": 0.9275224367691053,
      "grad_norm": 0.345703125,
      "learning_rate": 3.1729309667265063e-07,
      "loss": 0.5574,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6821
    },
    {
      "epoch": 0.927658417187925,
      "grad_norm": 0.361328125,
      "learning_rate": 3.1610770517485066e-07,
      "loss": 0.6178,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6822
    },
    {
      "epoch": 0.9277943976067446,
      "grad_norm": 0.3515625,
      "learning_rate": 3.149244965447684e-07,
      "loss": 0.7627,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6823
    },
    {
      "epoch": 0.9279303780255643,
      "grad_norm": 0.380859375,
      "learning_rate": 3.1374347104911716e-07,
      "loss": 0.6903,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6824
    },
    {
      "epoch": 0.928066358444384,
      "grad_norm": 0.41796875,
      "learning_rate": 3.1256462895411066e-07,
      "loss": 0.6635,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6825
    },
    {
      "epoch": 0.9282023388632037,
      "grad_norm": 0.32421875,
      "learning_rate": 3.113879705254774e-07,
      "loss": 0.6706,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6826
    },
    {
      "epoch": 0.9283383192820234,
      "grad_norm": 0.43359375,
      "learning_rate": 3.1021349602844755e-07,
      "loss": 0.7665,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6827
    },
    {
      "epoch": 0.928474299700843,
      "grad_norm": 0.390625,
      "learning_rate": 3.0904120572776583e-07,
      "loss": 0.6875,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6828
    },
    {
      "epoch": 0.9286102801196627,
      "grad_norm": 0.39453125,
      "learning_rate": 3.078710998876777e-07,
      "loss": 0.8224,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6829
    },
    {
      "epoch": 0.9287462605384824,
      "grad_norm": 0.5390625,
      "learning_rate": 3.06703178771941e-07,
      "loss": 0.5381,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6830
    },
    {
      "epoch": 0.9288822409573021,
      "grad_norm": 0.5234375,
      "learning_rate": 3.055374426438185e-07,
      "loss": 0.6224,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6831
    },
    {
      "epoch": 0.9290182213761219,
      "grad_norm": 0.37890625,
      "learning_rate": 3.043738917660821e-07,
      "loss": 0.6584,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6832
    },
    {
      "epoch": 0.9291542017949416,
      "grad_norm": 0.3515625,
      "learning_rate": 3.0321252640100883e-07,
      "loss": 0.6308,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6833
    },
    {
      "epoch": 0.9292901822137613,
      "grad_norm": 0.376953125,
      "learning_rate": 3.02053346810387e-07,
      "loss": 0.6764,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6834
    },
    {
      "epoch": 0.9294261626325809,
      "grad_norm": 0.455078125,
      "learning_rate": 3.008963532555076e-07,
      "loss": 0.8034,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6835
    },
    {
      "epoch": 0.9295621430514006,
      "grad_norm": 1.0625,
      "learning_rate": 2.997415459971731e-07,
      "loss": 0.9098,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6836
    },
    {
      "epoch": 0.9296981234702203,
      "grad_norm": 0.3125,
      "learning_rate": 2.985889252956897e-07,
      "loss": 0.6838,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6837
    },
    {
      "epoch": 0.92983410388904,
      "grad_norm": 0.54296875,
      "learning_rate": 2.97438491410873e-07,
      "loss": 0.7303,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6838
    },
    {
      "epoch": 0.9299700843078597,
      "grad_norm": 0.59765625,
      "learning_rate": 2.9629024460204434e-07,
      "loss": 0.6795,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6839
    },
    {
      "epoch": 0.9301060647266793,
      "grad_norm": 0.380859375,
      "learning_rate": 2.9514418512803345e-07,
      "loss": 0.6403,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6840
    },
    {
      "epoch": 0.930242045145499,
      "grad_norm": 0.310546875,
      "learning_rate": 2.9400031324717473e-07,
      "loss": 0.5684,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6841
    },
    {
      "epoch": 0.9303780255643187,
      "grad_norm": 0.375,
      "learning_rate": 2.928586292173119e-07,
      "loss": 0.6196,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6842
    },
    {
      "epoch": 0.9305140059831384,
      "grad_norm": 0.365234375,
      "learning_rate": 2.917191332957958e-07,
      "loss": 0.6774,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6843
    },
    {
      "epoch": 0.9306499864019581,
      "grad_norm": 0.22265625,
      "learning_rate": 2.905818257394799e-07,
      "loss": 0.4183,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6844
    },
    {
      "epoch": 0.9307859668207779,
      "grad_norm": 0.40234375,
      "learning_rate": 2.894467068047291e-07,
      "loss": 0.7749,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6845
    },
    {
      "epoch": 0.9309219472395975,
      "grad_norm": 0.53125,
      "learning_rate": 2.8831377674741203e-07,
      "loss": 0.6089,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6846
    },
    {
      "epoch": 0.9310579276584172,
      "grad_norm": 0.380859375,
      "learning_rate": 2.8718303582290665e-07,
      "loss": 0.7057,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6847
    },
    {
      "epoch": 0.9311939080772369,
      "grad_norm": 0.48046875,
      "learning_rate": 2.860544842860924e-07,
      "loss": 0.6484,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6848
    },
    {
      "epoch": 0.9313298884960566,
      "grad_norm": 0.53515625,
      "learning_rate": 2.849281223913636e-07,
      "loss": 0.7813,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6849
    },
    {
      "epoch": 0.9314658689148763,
      "grad_norm": 0.3203125,
      "learning_rate": 2.8380395039261155e-07,
      "loss": 0.5119,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6850
    },
    {
      "epoch": 0.9316018493336959,
      "grad_norm": 0.53515625,
      "learning_rate": 2.826819685432414e-07,
      "loss": 0.7531,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6851
    },
    {
      "epoch": 0.9317378297525156,
      "grad_norm": 0.3046875,
      "learning_rate": 2.815621770961585e-07,
      "loss": 0.5293,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6852
    },
    {
      "epoch": 0.9318738101713353,
      "grad_norm": 0.609375,
      "learning_rate": 2.804445763037811e-07,
      "loss": 0.5591,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6853
    },
    {
      "epoch": 0.932009790590155,
      "grad_norm": 0.4453125,
      "learning_rate": 2.793291664180264e-07,
      "loss": 0.6396,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6854
    },
    {
      "epoch": 0.9321457710089747,
      "grad_norm": 0.361328125,
      "learning_rate": 2.782159476903246e-07,
      "loss": 0.6878,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6855
    },
    {
      "epoch": 0.9322817514277943,
      "grad_norm": 0.427734375,
      "learning_rate": 2.77104920371607e-07,
      "loss": 0.5461,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6856
    },
    {
      "epoch": 0.9324177318466141,
      "grad_norm": 0.3515625,
      "learning_rate": 2.7599608471231333e-07,
      "loss": 0.7122,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6857
    },
    {
      "epoch": 0.9325537122654338,
      "grad_norm": 0.375,
      "learning_rate": 2.748894409623881e-07,
      "loss": 0.7946,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6858
    },
    {
      "epoch": 0.9326896926842535,
      "grad_norm": 0.89453125,
      "learning_rate": 2.737849893712841e-07,
      "loss": 0.6182,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6859
    },
    {
      "epoch": 0.9328256731030732,
      "grad_norm": 0.75390625,
      "learning_rate": 2.726827301879553e-07,
      "loss": 0.7819,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6860
    },
    {
      "epoch": 0.9329616535218929,
      "grad_norm": 0.37890625,
      "learning_rate": 2.715826636608676e-07,
      "loss": 0.6854,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6861
    },
    {
      "epoch": 0.9330976339407125,
      "grad_norm": 0.34765625,
      "learning_rate": 2.70484790037987e-07,
      "loss": 0.6226,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6862
    },
    {
      "epoch": 0.9332336143595322,
      "grad_norm": 0.404296875,
      "learning_rate": 2.6938910956679e-07,
      "loss": 0.606,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6863
    },
    {
      "epoch": 0.9333695947783519,
      "grad_norm": 0.451171875,
      "learning_rate": 2.6829562249425234e-07,
      "loss": 0.7562,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6864
    },
    {
      "epoch": 0.9335055751971716,
      "grad_norm": 0.47265625,
      "learning_rate": 2.6720432906686467e-07,
      "loss": 0.7599,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6865
    },
    {
      "epoch": 0.9336415556159913,
      "grad_norm": 0.2431640625,
      "learning_rate": 2.6611522953061353e-07,
      "loss": 0.4733,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6866
    },
    {
      "epoch": 0.933777536034811,
      "grad_norm": 0.490234375,
      "learning_rate": 2.65028324130997e-07,
      "loss": 0.6641,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6867
    },
    {
      "epoch": 0.9339135164536306,
      "grad_norm": 0.435546875,
      "learning_rate": 2.6394361311301685e-07,
      "loss": 0.5837,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6868
    },
    {
      "epoch": 0.9340494968724503,
      "grad_norm": 0.41796875,
      "learning_rate": 2.6286109672118086e-07,
      "loss": 0.5625,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6869
    },
    {
      "epoch": 0.9341854772912701,
      "grad_norm": 0.828125,
      "learning_rate": 2.617807751994994e-07,
      "loss": 0.6877,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6870
    },
    {
      "epoch": 0.9343214577100898,
      "grad_norm": 0.6484375,
      "learning_rate": 2.6070264879149323e-07,
      "loss": 0.9129,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6871
    },
    {
      "epoch": 0.9344574381289095,
      "grad_norm": 0.361328125,
      "learning_rate": 2.5962671774018234e-07,
      "loss": 0.6271,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6872
    },
    {
      "epoch": 0.9345934185477291,
      "grad_norm": 0.56640625,
      "learning_rate": 2.585529822880961e-07,
      "loss": 0.7729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6873
    },
    {
      "epoch": 0.9347293989665488,
      "grad_norm": 0.5078125,
      "learning_rate": 2.5748144267726646e-07,
      "loss": 0.8302,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6874
    },
    {
      "epoch": 0.9348653793853685,
      "grad_norm": 0.5390625,
      "learning_rate": 2.564120991492336e-07,
      "loss": 0.6074,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6875
    },
    {
      "epoch": 0.9350013598041882,
      "grad_norm": 0.39453125,
      "learning_rate": 2.553449519450391e-07,
      "loss": 0.6216,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6876
    },
    {
      "epoch": 0.9351373402230079,
      "grad_norm": 0.427734375,
      "learning_rate": 2.542800013052316e-07,
      "loss": 0.5523,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6877
    },
    {
      "epoch": 0.9352733206418276,
      "grad_norm": 0.625,
      "learning_rate": 2.532172474698624e-07,
      "loss": 0.8509,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6878
    },
    {
      "epoch": 0.9354093010606472,
      "grad_norm": 0.578125,
      "learning_rate": 2.5215669067849225e-07,
      "loss": 0.805,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6879
    },
    {
      "epoch": 0.9355452814794669,
      "grad_norm": 0.375,
      "learning_rate": 2.510983311701809e-07,
      "loss": 0.542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6880
    },
    {
      "epoch": 0.9356812618982866,
      "grad_norm": 0.419921875,
      "learning_rate": 2.500421691834975e-07,
      "loss": 0.6833,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6881
    },
    {
      "epoch": 0.9358172423171064,
      "grad_norm": 0.94140625,
      "learning_rate": 2.4898820495651175e-07,
      "loss": 0.6864,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6882
    },
    {
      "epoch": 0.9359532227359261,
      "grad_norm": 0.384765625,
      "learning_rate": 2.479364387268024e-07,
      "loss": 0.6918,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6883
    },
    {
      "epoch": 0.9360892031547458,
      "grad_norm": 0.373046875,
      "learning_rate": 2.4688687073144867e-07,
      "loss": 0.5794,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6884
    },
    {
      "epoch": 0.9362251835735654,
      "grad_norm": 0.5546875,
      "learning_rate": 2.458395012070369e-07,
      "loss": 0.6424,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6885
    },
    {
      "epoch": 0.9363611639923851,
      "grad_norm": 0.44140625,
      "learning_rate": 2.4479433038965606e-07,
      "loss": 0.7687,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6886
    },
    {
      "epoch": 0.9364971444112048,
      "grad_norm": 0.3046875,
      "learning_rate": 2.4375135851490206e-07,
      "loss": 0.5688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6887
    },
    {
      "epoch": 0.9366331248300245,
      "grad_norm": 0.333984375,
      "learning_rate": 2.4271058581787245e-07,
      "loss": 0.5208,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6888
    },
    {
      "epoch": 0.9367691052488442,
      "grad_norm": 0.5,
      "learning_rate": 2.416720125331695e-07,
      "loss": 0.8267,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6889
    },
    {
      "epoch": 0.9369050856676638,
      "grad_norm": 0.28515625,
      "learning_rate": 2.406356388949005e-07,
      "loss": 0.5288,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6890
    },
    {
      "epoch": 0.9370410660864835,
      "grad_norm": 0.37890625,
      "learning_rate": 2.396014651366785e-07,
      "loss": 0.6414,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6891
    },
    {
      "epoch": 0.9371770465053032,
      "grad_norm": 0.318359375,
      "learning_rate": 2.3856949149161703e-07,
      "loss": 0.641,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6892
    },
    {
      "epoch": 0.9373130269241229,
      "grad_norm": 0.30859375,
      "learning_rate": 2.3753971819233668e-07,
      "loss": 0.61,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6893
    },
    {
      "epoch": 0.9374490073429426,
      "grad_norm": 0.9375,
      "learning_rate": 2.3651214547095957e-07,
      "loss": 0.6105,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6894
    },
    {
      "epoch": 0.9375849877617624,
      "grad_norm": 0.412109375,
      "learning_rate": 2.3548677355911599e-07,
      "loss": 0.5708,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6895
    },
    {
      "epoch": 0.937720968180582,
      "grad_norm": 0.38671875,
      "learning_rate": 2.3446360268793323e-07,
      "loss": 0.7077,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6896
    },
    {
      "epoch": 0.9378569485994017,
      "grad_norm": 0.380859375,
      "learning_rate": 2.3344263308805016e-07,
      "loss": 0.7332,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6897
    },
    {
      "epoch": 0.9379929290182214,
      "grad_norm": 0.49609375,
      "learning_rate": 2.3242386498960267e-07,
      "loss": 0.5915,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6898
    },
    {
      "epoch": 0.9381289094370411,
      "grad_norm": 0.47265625,
      "learning_rate": 2.3140729862223599e-07,
      "loss": 0.6257,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6899
    },
    {
      "epoch": 0.9382648898558608,
      "grad_norm": 0.41015625,
      "learning_rate": 2.3039293421509567e-07,
      "loss": 0.6139,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6900
    },
    {
      "epoch": 0.9384008702746804,
      "grad_norm": 0.52734375,
      "learning_rate": 2.293807719968333e-07,
      "loss": 0.7741,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6901
    },
    {
      "epoch": 0.9385368506935001,
      "grad_norm": 0.7578125,
      "learning_rate": 2.2837081219559963e-07,
      "loss": 0.7879,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6902
    },
    {
      "epoch": 0.9386728311123198,
      "grad_norm": 0.63671875,
      "learning_rate": 2.2736305503905376e-07,
      "loss": 0.8618,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6903
    },
    {
      "epoch": 0.9388088115311395,
      "grad_norm": 0.2353515625,
      "learning_rate": 2.2635750075435725e-07,
      "loss": 0.5103,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6904
    },
    {
      "epoch": 0.9389447919499592,
      "grad_norm": 0.333984375,
      "learning_rate": 2.2535414956817213e-07,
      "loss": 0.6053,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6905
    },
    {
      "epoch": 0.9390807723687788,
      "grad_norm": 0.44140625,
      "learning_rate": 2.2435300170666863e-07,
      "loss": 0.415,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6906
    },
    {
      "epoch": 0.9392167527875986,
      "grad_norm": 0.2392578125,
      "learning_rate": 2.2335405739551507e-07,
      "loss": 0.4313,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6907
    },
    {
      "epoch": 0.9393527332064183,
      "grad_norm": 0.36328125,
      "learning_rate": 2.2235731685988804e-07,
      "loss": 0.579,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6908
    },
    {
      "epoch": 0.939488713625238,
      "grad_norm": 0.248046875,
      "learning_rate": 2.2136278032446335e-07,
      "loss": 0.5008,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6909
    },
    {
      "epoch": 0.9396246940440577,
      "grad_norm": 0.232421875,
      "learning_rate": 2.203704480134228e-07,
      "loss": 0.4043,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6910
    },
    {
      "epoch": 0.9397606744628774,
      "grad_norm": 0.291015625,
      "learning_rate": 2.1938032015044964e-07,
      "loss": 0.5308,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6911
    },
    {
      "epoch": 0.939896654881697,
      "grad_norm": 0.68359375,
      "learning_rate": 2.1839239695873205e-07,
      "loss": 0.7614,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6912
    },
    {
      "epoch": 0.9400326353005167,
      "grad_norm": 0.34765625,
      "learning_rate": 2.1740667866095744e-07,
      "loss": 0.451,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6913
    },
    {
      "epoch": 0.9401686157193364,
      "grad_norm": 0.59765625,
      "learning_rate": 2.164231654793203e-07,
      "loss": 0.5835,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6914
    },
    {
      "epoch": 0.9403045961381561,
      "grad_norm": 0.421875,
      "learning_rate": 2.154418576355166e-07,
      "loss": 0.8828,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6915
    },
    {
      "epoch": 0.9404405765569758,
      "grad_norm": 0.365234375,
      "learning_rate": 2.1446275535074612e-07,
      "loss": 0.6857,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6916
    },
    {
      "epoch": 0.9405765569757955,
      "grad_norm": 0.4609375,
      "learning_rate": 2.1348585884570781e-07,
      "loss": 0.8372,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6917
    },
    {
      "epoch": 0.9407125373946151,
      "grad_norm": 0.5,
      "learning_rate": 2.1251116834060893e-07,
      "loss": 0.583,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6918
    },
    {
      "epoch": 0.9408485178134348,
      "grad_norm": 0.333984375,
      "learning_rate": 2.1153868405515364e-07,
      "loss": 0.6331,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6919
    },
    {
      "epoch": 0.9409844982322546,
      "grad_norm": 0.470703125,
      "learning_rate": 2.1056840620855555e-07,
      "loss": 0.4736,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6920
    },
    {
      "epoch": 0.9411204786510743,
      "grad_norm": 0.39453125,
      "learning_rate": 2.0960033501952305e-07,
      "loss": 0.61,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6921
    },
    {
      "epoch": 0.941256459069894,
      "grad_norm": 0.27734375,
      "learning_rate": 2.0863447070627484e-07,
      "loss": 0.529,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6922
    },
    {
      "epoch": 0.9413924394887137,
      "grad_norm": 0.46484375,
      "learning_rate": 2.0767081348652574e-07,
      "loss": 0.7189,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6923
    },
    {
      "epoch": 0.9415284199075333,
      "grad_norm": 2.328125,
      "learning_rate": 2.067093635774975e-07,
      "loss": 0.6426,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6924
    },
    {
      "epoch": 0.941664400326353,
      "grad_norm": 0.486328125,
      "learning_rate": 2.057501211959123e-07,
      "loss": 0.7865,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6925
    },
    {
      "epoch": 0.9418003807451727,
      "grad_norm": 0.31640625,
      "learning_rate": 2.0479308655799613e-07,
      "loss": 0.6502,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6926
    },
    {
      "epoch": 0.9419363611639924,
      "grad_norm": 0.31640625,
      "learning_rate": 2.0383825987947303e-07,
      "loss": 0.5623,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6927
    },
    {
      "epoch": 0.9420723415828121,
      "grad_norm": 0.609375,
      "learning_rate": 2.0288564137557642e-07,
      "loss": 0.6729,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6928
    },
    {
      "epoch": 0.9422083220016317,
      "grad_norm": 0.30078125,
      "learning_rate": 2.0193523126103565e-07,
      "loss": 0.6432,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6929
    },
    {
      "epoch": 0.9423443024204514,
      "grad_norm": 0.416015625,
      "learning_rate": 2.009870297500871e-07,
      "loss": 0.7463,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6930
    },
    {
      "epoch": 0.9424802828392711,
      "grad_norm": 0.34375,
      "learning_rate": 2.0004103705646426e-07,
      "loss": 0.5472,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6931
    },
    {
      "epoch": 0.9426162632580908,
      "grad_norm": 0.45703125,
      "learning_rate": 1.990972533934077e-07,
      "loss": 0.6556,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6932
    },
    {
      "epoch": 0.9427522436769106,
      "grad_norm": 0.423828125,
      "learning_rate": 1.9815567897365608e-07,
      "loss": 0.7738,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6933
    },
    {
      "epoch": 0.9428882240957303,
      "grad_norm": 0.341796875,
      "learning_rate": 1.9721631400945295e-07,
      "loss": 0.7443,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6934
    },
    {
      "epoch": 0.9430242045145499,
      "grad_norm": 0.408203125,
      "learning_rate": 1.962791587125412e-07,
      "loss": 0.5549,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6935
    },
    {
      "epoch": 0.9431601849333696,
      "grad_norm": 0.328125,
      "learning_rate": 1.9534421329416963e-07,
      "loss": 0.6118,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6936
    },
    {
      "epoch": 0.9432961653521893,
      "grad_norm": 0.38671875,
      "learning_rate": 1.9441147796508408e-07,
      "loss": 0.6895,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6937
    },
    {
      "epoch": 0.943432145771009,
      "grad_norm": 0.609375,
      "learning_rate": 1.9348095293553526e-07,
      "loss": 0.4818,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6938
    },
    {
      "epoch": 0.9435681261898287,
      "grad_norm": 0.35546875,
      "learning_rate": 1.9255263841527537e-07,
      "loss": 0.6744,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6939
    },
    {
      "epoch": 0.9437041066086483,
      "grad_norm": 0.341796875,
      "learning_rate": 1.9162653461355708e-07,
      "loss": 0.5866,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6940
    },
    {
      "epoch": 0.943840087027468,
      "grad_norm": 0.8125,
      "learning_rate": 1.9070264173913554e-07,
      "loss": 0.7401,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6941
    },
    {
      "epoch": 0.9439760674462877,
      "grad_norm": 0.484375,
      "learning_rate": 1.8978096000026868e-07,
      "loss": 0.7576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6942
    },
    {
      "epoch": 0.9441120478651074,
      "grad_norm": 0.380859375,
      "learning_rate": 1.8886148960471362e-07,
      "loss": 0.4884,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6943
    },
    {
      "epoch": 0.9442480282839271,
      "grad_norm": 0.3515625,
      "learning_rate": 1.8794423075973013e-07,
      "loss": 0.5721,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6944
    },
    {
      "epoch": 0.9443840087027469,
      "grad_norm": 0.373046875,
      "learning_rate": 1.8702918367208057e-07,
      "loss": 0.6681,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6945
    },
    {
      "epoch": 0.9445199891215665,
      "grad_norm": 0.30078125,
      "learning_rate": 1.8611634854802663e-07,
      "loss": 0.4727,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6946
    },
    {
      "epoch": 0.9446559695403862,
      "grad_norm": 0.8203125,
      "learning_rate": 1.8520572559333372e-07,
      "loss": 0.9168,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6947
    },
    {
      "epoch": 0.9447919499592059,
      "grad_norm": 0.34765625,
      "learning_rate": 1.8429731501326652e-07,
      "loss": 0.7495,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6948
    },
    {
      "epoch": 0.9449279303780256,
      "grad_norm": 0.330078125,
      "learning_rate": 1.8339111701259126e-07,
      "loss": 0.4784,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6949
    },
    {
      "epoch": 0.9450639107968453,
      "grad_norm": 0.484375,
      "learning_rate": 1.8248713179557788e-07,
      "loss": 0.8898,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6950
    },
    {
      "epoch": 0.945199891215665,
      "grad_norm": 0.3515625,
      "learning_rate": 1.8158535956599332e-07,
      "loss": 0.569,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6951
    },
    {
      "epoch": 0.9453358716344846,
      "grad_norm": 0.43359375,
      "learning_rate": 1.8068580052711172e-07,
      "loss": 0.526,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6952
    },
    {
      "epoch": 0.9454718520533043,
      "grad_norm": 0.380859375,
      "learning_rate": 1.797884548816997e-07,
      "loss": 0.7731,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6953
    },
    {
      "epoch": 0.945607832472124,
      "grad_norm": 0.77734375,
      "learning_rate": 1.7889332283203442e-07,
      "loss": 0.752,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6954
    },
    {
      "epoch": 0.9457438128909437,
      "grad_norm": 0.349609375,
      "learning_rate": 1.7800040457988777e-07,
      "loss": 0.6082,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6955
    },
    {
      "epoch": 0.9458797933097634,
      "grad_norm": 0.40625,
      "learning_rate": 1.771097003265343e-07,
      "loss": 0.7525,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6956
    },
    {
      "epoch": 0.946015773728583,
      "grad_norm": 0.490234375,
      "learning_rate": 1.7622121027274897e-07,
      "loss": 0.6268,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6957
    },
    {
      "epoch": 0.9461517541474028,
      "grad_norm": 0.37109375,
      "learning_rate": 1.7533493461881045e-07,
      "loss": 0.832,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6958
    },
    {
      "epoch": 0.9462877345662225,
      "grad_norm": 0.671875,
      "learning_rate": 1.7445087356449343e-07,
      "loss": 0.8389,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6959
    },
    {
      "epoch": 0.9464237149850422,
      "grad_norm": 0.333984375,
      "learning_rate": 1.7356902730907843e-07,
      "loss": 0.6484,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6960
    },
    {
      "epoch": 0.9465596954038619,
      "grad_norm": 0.68359375,
      "learning_rate": 1.7268939605134317e-07,
      "loss": 0.5396,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6961
    },
    {
      "epoch": 0.9466956758226815,
      "grad_norm": 0.328125,
      "learning_rate": 1.718119799895668e-07,
      "loss": 0.6548,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6962
    },
    {
      "epoch": 0.9468316562415012,
      "grad_norm": 0.255859375,
      "learning_rate": 1.7093677932153218e-07,
      "loss": 0.5524,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6963
    },
    {
      "epoch": 0.9469676366603209,
      "grad_norm": 0.478515625,
      "learning_rate": 1.7006379424451602e-07,
      "loss": 0.7222,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6964
    },
    {
      "epoch": 0.9471036170791406,
      "grad_norm": 1.6796875,
      "learning_rate": 1.691930249553042e-07,
      "loss": 0.8695,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6965
    },
    {
      "epoch": 0.9472395974979603,
      "grad_norm": 0.6015625,
      "learning_rate": 1.683244716501764e-07,
      "loss": 0.6372,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6966
    },
    {
      "epoch": 0.94737557791678,
      "grad_norm": 0.294921875,
      "learning_rate": 1.674581345249149e-07,
      "loss": 0.7409,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6967
    },
    {
      "epoch": 0.9475115583355996,
      "grad_norm": 0.45703125,
      "learning_rate": 1.6659401377480344e-07,
      "loss": 0.6475,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6968
    },
    {
      "epoch": 0.9476475387544193,
      "grad_norm": 0.345703125,
      "learning_rate": 1.6573210959462628e-07,
      "loss": 0.7743,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6969
    },
    {
      "epoch": 0.9477835191732391,
      "grad_norm": 0.21875,
      "learning_rate": 1.6487242217866462e-07,
      "loss": 0.4255,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6970
    },
    {
      "epoch": 0.9479194995920588,
      "grad_norm": 0.443359375,
      "learning_rate": 1.640149517207057e-07,
      "loss": 0.7546,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6971
    },
    {
      "epoch": 0.9480554800108785,
      "grad_norm": 0.23828125,
      "learning_rate": 1.6315969841403045e-07,
      "loss": 0.4214,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6972
    },
    {
      "epoch": 0.9481914604296982,
      "grad_norm": 0.55078125,
      "learning_rate": 1.623066624514258e-07,
      "loss": 0.861,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6973
    },
    {
      "epoch": 0.9483274408485178,
      "grad_norm": 0.419921875,
      "learning_rate": 1.6145584402517568e-07,
      "loss": 0.5876,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6974
    },
    {
      "epoch": 0.9484634212673375,
      "grad_norm": 0.51171875,
      "learning_rate": 1.6060724332706667e-07,
      "loss": 0.8271,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6975
    },
    {
      "epoch": 0.9485994016861572,
      "grad_norm": 0.6171875,
      "learning_rate": 1.5976086054838024e-07,
      "loss": 0.6273,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6976
    },
    {
      "epoch": 0.9487353821049769,
      "grad_norm": 0.34375,
      "learning_rate": 1.589166958799049e-07,
      "loss": 0.5422,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6977
    },
    {
      "epoch": 0.9488713625237966,
      "grad_norm": 0.443359375,
      "learning_rate": 1.5807474951192282e-07,
      "loss": 0.793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6978
    },
    {
      "epoch": 0.9490073429426162,
      "grad_norm": 0.345703125,
      "learning_rate": 1.572350216342211e-07,
      "loss": 0.592,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6979
    },
    {
      "epoch": 0.9491433233614359,
      "grad_norm": 0.375,
      "learning_rate": 1.5639751243608393e-07,
      "loss": 0.6519,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6980
    },
    {
      "epoch": 0.9492793037802556,
      "grad_norm": 0.68359375,
      "learning_rate": 1.555622221062969e-07,
      "loss": 0.6283,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6981
    },
    {
      "epoch": 0.9494152841990753,
      "grad_norm": 0.40625,
      "learning_rate": 1.5472915083314277e-07,
      "loss": 0.7985,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6982
    },
    {
      "epoch": 0.9495512646178951,
      "grad_norm": 0.37109375,
      "learning_rate": 1.5389829880440688e-07,
      "loss": 0.652,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6983
    },
    {
      "epoch": 0.9496872450367148,
      "grad_norm": 0.40234375,
      "learning_rate": 1.5306966620737385e-07,
      "loss": 0.7985,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6984
    },
    {
      "epoch": 0.9498232254555344,
      "grad_norm": 0.259765625,
      "learning_rate": 1.5224325322882872e-07,
      "loss": 0.4233,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6985
    },
    {
      "epoch": 0.9499592058743541,
      "grad_norm": 0.30859375,
      "learning_rate": 1.5141906005505248e-07,
      "loss": 0.527,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6986
    },
    {
      "epoch": 0.9500951862931738,
      "grad_norm": 0.478515625,
      "learning_rate": 1.5059708687182984e-07,
      "loss": 0.8356,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6987
    },
    {
      "epoch": 0.9502311667119935,
      "grad_norm": 0.28515625,
      "learning_rate": 1.497773338644437e-07,
      "loss": 0.584,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6988
    },
    {
      "epoch": 0.9503671471308132,
      "grad_norm": 0.322265625,
      "learning_rate": 1.4895980121767627e-07,
      "loss": 0.641,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6989
    },
    {
      "epoch": 0.9505031275496328,
      "grad_norm": 0.318359375,
      "learning_rate": 1.481444891158079e-07,
      "loss": 0.665,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6990
    },
    {
      "epoch": 0.9506391079684525,
      "grad_norm": 0.28125,
      "learning_rate": 1.4733139774262274e-07,
      "loss": 0.5047,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6991
    },
    {
      "epoch": 0.9507750883872722,
      "grad_norm": 0.3828125,
      "learning_rate": 1.4652052728139964e-07,
      "loss": 0.752,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6992
    },
    {
      "epoch": 0.9509110688060919,
      "grad_norm": 0.369140625,
      "learning_rate": 1.4571187791491915e-07,
      "loss": 0.6255,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6993
    },
    {
      "epoch": 0.9510470492249116,
      "grad_norm": 0.359375,
      "learning_rate": 1.4490544982545984e-07,
      "loss": 0.7511,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6994
    },
    {
      "epoch": 0.9511830296437312,
      "grad_norm": 0.326171875,
      "learning_rate": 1.44101243194803e-07,
      "loss": 0.6489,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6995
    },
    {
      "epoch": 0.951319010062551,
      "grad_norm": 0.404296875,
      "learning_rate": 1.4329925820422364e-07,
      "loss": 0.7347,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6996
    },
    {
      "epoch": 0.9514549904813707,
      "grad_norm": 0.306640625,
      "learning_rate": 1.4249949503450045e-07,
      "loss": 0.5942,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6997
    },
    {
      "epoch": 0.9515909709001904,
      "grad_norm": 0.43359375,
      "learning_rate": 1.4170195386590923e-07,
      "loss": 0.5054,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6998
    },
    {
      "epoch": 0.9517269513190101,
      "grad_norm": 0.6796875,
      "learning_rate": 1.4090663487822732e-07,
      "loss": 0.5685,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 6999
    },
    {
      "epoch": 0.9518629317378298,
      "grad_norm": 0.390625,
      "learning_rate": 1.4011353825072683e-07,
      "loss": 0.741,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7000
    },
    {
      "epoch": 0.9519989121566494,
      "grad_norm": 0.52734375,
      "learning_rate": 1.3932266416218366e-07,
      "loss": 0.696,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7001
    },
    {
      "epoch": 0.9521348925754691,
      "grad_norm": 0.57421875,
      "learning_rate": 1.3853401279086853e-07,
      "loss": 0.8178,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7002
    },
    {
      "epoch": 0.9522708729942888,
      "grad_norm": 0.5078125,
      "learning_rate": 1.377475843145537e-07,
      "loss": 0.4191,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7003
    },
    {
      "epoch": 0.9524068534131085,
      "grad_norm": 0.9453125,
      "learning_rate": 1.369633789105107e-07,
      "loss": 0.5112,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7004
    },
    {
      "epoch": 0.9525428338319282,
      "grad_norm": 0.259765625,
      "learning_rate": 1.3618139675550813e-07,
      "loss": 0.4824,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7005
    },
    {
      "epoch": 0.9526788142507479,
      "grad_norm": 0.57421875,
      "learning_rate": 1.3540163802581496e-07,
      "loss": 0.6286,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7006
    },
    {
      "epoch": 0.9528147946695675,
      "grad_norm": 0.45703125,
      "learning_rate": 1.3462410289719728e-07,
      "loss": 0.5938,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7007
    },
    {
      "epoch": 0.9529507750883873,
      "grad_norm": 0.408203125,
      "learning_rate": 1.3384879154492158e-07,
      "loss": 0.6125,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7008
    },
    {
      "epoch": 0.953086755507207,
      "grad_norm": 0.55078125,
      "learning_rate": 1.330757041437536e-07,
      "loss": 0.6279,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7009
    },
    {
      "epoch": 0.9532227359260267,
      "grad_norm": 1.921875,
      "learning_rate": 1.3230484086795503e-07,
      "loss": 0.7609,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7010
    },
    {
      "epoch": 0.9533587163448464,
      "grad_norm": 0.462890625,
      "learning_rate": 1.3153620189128913e-07,
      "loss": 0.835,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7011
    },
    {
      "epoch": 0.953494696763666,
      "grad_norm": 0.42578125,
      "learning_rate": 1.3076978738701507e-07,
      "loss": 0.7088,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7012
    },
    {
      "epoch": 0.9536306771824857,
      "grad_norm": 0.296875,
      "learning_rate": 1.3000559752789354e-07,
      "loss": 0.6012,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7013
    },
    {
      "epoch": 0.9537666576013054,
      "grad_norm": 0.20703125,
      "learning_rate": 1.2924363248618122e-07,
      "loss": 0.4287,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7014
    },
    {
      "epoch": 0.9539026380201251,
      "grad_norm": 0.59765625,
      "learning_rate": 1.2848389243363514e-07,
      "loss": 0.5016,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7015
    },
    {
      "epoch": 0.9540386184389448,
      "grad_norm": 0.6015625,
      "learning_rate": 1.2772637754150941e-07,
      "loss": 0.495,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7016
    },
    {
      "epoch": 0.9541745988577645,
      "grad_norm": 0.3125,
      "learning_rate": 1.2697108798055745e-07,
      "loss": 0.6499,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7017
    },
    {
      "epoch": 0.9543105792765841,
      "grad_norm": 0.306640625,
      "learning_rate": 1.262180239210309e-07,
      "loss": 0.5376,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7018
    },
    {
      "epoch": 0.9544465596954038,
      "grad_norm": 0.640625,
      "learning_rate": 1.2546718553267944e-07,
      "loss": 0.6836,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7019
    },
    {
      "epoch": 0.9545825401142235,
      "grad_norm": 0.3359375,
      "learning_rate": 1.2471857298475108e-07,
      "loss": 0.6908,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7020
    },
    {
      "epoch": 0.9547185205330433,
      "grad_norm": 0.6015625,
      "learning_rate": 1.2397218644599197e-07,
      "loss": 0.7731,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7021
    },
    {
      "epoch": 0.954854500951863,
      "grad_norm": 0.474609375,
      "learning_rate": 1.2322802608464746e-07,
      "loss": 0.7031,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7022
    },
    {
      "epoch": 0.9549904813706827,
      "grad_norm": 0.39453125,
      "learning_rate": 1.2248609206846008e-07,
      "loss": 0.7008,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7023
    },
    {
      "epoch": 0.9551264617895023,
      "grad_norm": 0.28515625,
      "learning_rate": 1.2174638456467046e-07,
      "loss": 0.5412,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7024
    },
    {
      "epoch": 0.955262442208322,
      "grad_norm": 0.35546875,
      "learning_rate": 1.210089037400186e-07,
      "loss": 0.5365,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7025
    },
    {
      "epoch": 0.9553984226271417,
      "grad_norm": 0.9921875,
      "learning_rate": 1.2027364976074152e-07,
      "loss": 0.9142,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7026
    },
    {
      "epoch": 0.9555344030459614,
      "grad_norm": 0.5625,
      "learning_rate": 1.1954062279257328e-07,
      "loss": 0.7563,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7027
    },
    {
      "epoch": 0.9556703834647811,
      "grad_norm": 0.337890625,
      "learning_rate": 1.1880982300074839e-07,
      "loss": 0.7264,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7028
    },
    {
      "epoch": 0.9558063638836007,
      "grad_norm": 0.451171875,
      "learning_rate": 1.180812505499973e-07,
      "loss": 0.7448,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7029
    },
    {
      "epoch": 0.9559423443024204,
      "grad_norm": 0.302734375,
      "learning_rate": 1.1735490560454976e-07,
      "loss": 0.5132,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7030
    },
    {
      "epoch": 0.9560783247212401,
      "grad_norm": 0.69921875,
      "learning_rate": 1.1663078832813146e-07,
      "loss": 0.6971,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7031
    },
    {
      "epoch": 0.9562143051400598,
      "grad_norm": 0.181640625,
      "learning_rate": 1.1590889888396962e-07,
      "loss": 0.3283,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7032
    },
    {
      "epoch": 0.9563502855588796,
      "grad_norm": 0.53515625,
      "learning_rate": 1.1518923743478516e-07,
      "loss": 0.5511,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7033
    },
    {
      "epoch": 0.9564862659776993,
      "grad_norm": 0.259765625,
      "learning_rate": 1.1447180414279946e-07,
      "loss": 0.4567,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7034
    },
    {
      "epoch": 0.9566222463965189,
      "grad_norm": 0.45703125,
      "learning_rate": 1.137565991697298e-07,
      "loss": 0.5381,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7035
    },
    {
      "epoch": 0.9567582268153386,
      "grad_norm": 0.30078125,
      "learning_rate": 1.130436226767928e-07,
      "loss": 0.5542,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7036
    },
    {
      "epoch": 0.9568942072341583,
      "grad_norm": 0.57421875,
      "learning_rate": 1.1233287482470212e-07,
      "loss": 0.5578,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7037
    },
    {
      "epoch": 0.957030187652978,
      "grad_norm": 0.494140625,
      "learning_rate": 1.116243557736696e-07,
      "loss": 0.3091,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7038
    },
    {
      "epoch": 0.9571661680717977,
      "grad_norm": 0.54296875,
      "learning_rate": 1.109180656834019e-07,
      "loss": 0.668,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7039
    },
    {
      "epoch": 0.9573021484906173,
      "grad_norm": 0.34765625,
      "learning_rate": 1.102140047131084e-07,
      "loss": 0.5037,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7040
    },
    {
      "epoch": 0.957438128909437,
      "grad_norm": 0.7421875,
      "learning_rate": 1.0951217302148986e-07,
      "loss": 0.5964,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7041
    },
    {
      "epoch": 0.9575741093282567,
      "grad_norm": 0.2314453125,
      "learning_rate": 1.0881257076675088e-07,
      "loss": 0.4455,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7042
    },
    {
      "epoch": 0.9577100897470764,
      "grad_norm": 0.49609375,
      "learning_rate": 1.0811519810658866e-07,
      "loss": 0.5879,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7043
    },
    {
      "epoch": 0.9578460701658961,
      "grad_norm": 0.353515625,
      "learning_rate": 1.0742005519819964e-07,
      "loss": 0.665,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7044
    },
    {
      "epoch": 0.9579820505847157,
      "grad_norm": 0.390625,
      "learning_rate": 1.0672714219827629e-07,
      "loss": 0.721,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7045
    },
    {
      "epoch": 0.9581180310035355,
      "grad_norm": 0.177734375,
      "learning_rate": 1.0603645926301254e-07,
      "loss": 0.2944,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7046
    },
    {
      "epoch": 0.9582540114223552,
      "grad_norm": 0.40234375,
      "learning_rate": 1.0534800654809496e-07,
      "loss": 0.6499,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7047
    },
    {
      "epoch": 0.9583899918411749,
      "grad_norm": 0.3515625,
      "learning_rate": 1.0466178420870943e-07,
      "loss": 0.6336,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7048
    },
    {
      "epoch": 0.9585259722599946,
      "grad_norm": 0.431640625,
      "learning_rate": 1.0397779239953887e-07,
      "loss": 0.7622,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7049
    },
    {
      "epoch": 0.9586619526788143,
      "grad_norm": 0.396484375,
      "learning_rate": 1.0329603127476439e-07,
      "loss": 0.6463,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7050
    },
    {
      "epoch": 0.958797933097634,
      "grad_norm": 0.34765625,
      "learning_rate": 1.0261650098806086e-07,
      "loss": 0.7451,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7051
    },
    {
      "epoch": 0.9589339135164536,
      "grad_norm": 0.2578125,
      "learning_rate": 1.0193920169260574e-07,
      "loss": 0.5392,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7052
    },
    {
      "epoch": 0.9590698939352733,
      "grad_norm": 0.26171875,
      "learning_rate": 1.012641335410669e-07,
      "loss": 0.4539,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7053
    },
    {
      "epoch": 0.959205874354093,
      "grad_norm": 0.408203125,
      "learning_rate": 1.0059129668561707e-07,
      "loss": 0.597,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7054
    },
    {
      "epoch": 0.9593418547729127,
      "grad_norm": 0.30859375,
      "learning_rate": 9.992069127791937e-08,
      "loss": 0.543,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7055
    },
    {
      "epoch": 0.9594778351917324,
      "grad_norm": 0.486328125,
      "learning_rate": 9.925231746913622e-08,
      "loss": 0.7303,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7056
    },
    {
      "epoch": 0.959613815610552,
      "grad_norm": 0.322265625,
      "learning_rate": 9.858617540992932e-08,
      "loss": 0.5731,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7057
    },
    {
      "epoch": 0.9597497960293717,
      "grad_norm": 0.3359375,
      "learning_rate": 9.792226525045301e-08,
      "loss": 0.627,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7058
    },
    {
      "epoch": 0.9598857764481915,
      "grad_norm": 0.37109375,
      "learning_rate": 9.726058714036201e-08,
      "loss": 0.5093,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7059
    },
    {
      "epoch": 0.9600217568670112,
      "grad_norm": 0.3984375,
      "learning_rate": 9.660114122880593e-08,
      "loss": 0.5973,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7060
    },
    {
      "epoch": 0.9601577372858309,
      "grad_norm": 0.8828125,
      "learning_rate": 9.59439276644325e-08,
      "loss": 0.6021,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7061
    },
    {
      "epoch": 0.9602937177046506,
      "grad_norm": 0.34765625,
      "learning_rate": 9.528894659538657e-08,
      "loss": 0.6806,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7062
    },
    {
      "epoch": 0.9604296981234702,
      "grad_norm": 0.353515625,
      "learning_rate": 9.46361981693067e-08,
      "loss": 0.633,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7063
    },
    {
      "epoch": 0.9605656785422899,
      "grad_norm": 0.375,
      "learning_rate": 9.398568253333185e-08,
      "loss": 0.6802,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7064
    },
    {
      "epoch": 0.9607016589611096,
      "grad_norm": 0.72265625,
      "learning_rate": 9.333739983409584e-08,
      "loss": 0.7339,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7065
    },
    {
      "epoch": 0.9608376393799293,
      "grad_norm": 0.296875,
      "learning_rate": 9.269135021773068e-08,
      "loss": 0.5213,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7066
    },
    {
      "epoch": 0.960973619798749,
      "grad_norm": 0.640625,
      "learning_rate": 9.204753382986097e-08,
      "loss": 0.6225,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7067
    },
    {
      "epoch": 0.9611096002175686,
      "grad_norm": 0.404296875,
      "learning_rate": 9.140595081561399e-08,
      "loss": 0.6958,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7068
    },
    {
      "epoch": 0.9612455806363883,
      "grad_norm": 0.361328125,
      "learning_rate": 9.076660131960735e-08,
      "loss": 0.5936,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7069
    },
    {
      "epoch": 0.961381561055208,
      "grad_norm": 0.31640625,
      "learning_rate": 9.012948548596134e-08,
      "loss": 0.6383,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7070
    },
    {
      "epoch": 0.9615175414740278,
      "grad_norm": 0.345703125,
      "learning_rate": 8.949460345828776e-08,
      "loss": 0.6662,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7071
    },
    {
      "epoch": 0.9616535218928475,
      "grad_norm": 0.52734375,
      "learning_rate": 8.886195537969655e-08,
      "loss": 0.8561,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7072
    },
    {
      "epoch": 0.9617895023116672,
      "grad_norm": 0.451171875,
      "learning_rate": 8.823154139279588e-08,
      "loss": 0.7039,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7073
    },
    {
      "epoch": 0.9619254827304868,
      "grad_norm": 0.52734375,
      "learning_rate": 8.760336163968764e-08,
      "loss": 0.5137,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7074
    },
    {
      "epoch": 0.9620614631493065,
      "grad_norm": 0.3671875,
      "learning_rate": 8.697741626197076e-08,
      "loss": 0.5241,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7075
    },
    {
      "epoch": 0.9621974435681262,
      "grad_norm": 0.498046875,
      "learning_rate": 8.63537054007424e-08,
      "loss": 0.5682,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7076
    },
    {
      "epoch": 0.9623334239869459,
      "grad_norm": 0.3515625,
      "learning_rate": 8.573222919659341e-08,
      "loss": 0.6107,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7077
    },
    {
      "epoch": 0.9624694044057656,
      "grad_norm": 0.3359375,
      "learning_rate": 8.511298778961286e-08,
      "loss": 0.6883,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7078
    },
    {
      "epoch": 0.9626053848245852,
      "grad_norm": 0.8203125,
      "learning_rate": 8.449598131938464e-08,
      "loss": 0.7429,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7079
    },
    {
      "epoch": 0.9627413652434049,
      "grad_norm": 0.35546875,
      "learning_rate": 8.388120992499083e-08,
      "loss": 0.5654,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7080
    },
    {
      "epoch": 0.9628773456622246,
      "grad_norm": 0.4296875,
      "learning_rate": 8.326867374500613e-08,
      "loss": 0.8461,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7081
    },
    {
      "epoch": 0.9630133260810443,
      "grad_norm": 0.53125,
      "learning_rate": 8.265837291750677e-08,
      "loss": 0.7292,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7082
    },
    {
      "epoch": 0.963149306499864,
      "grad_norm": 0.19921875,
      "learning_rate": 8.205030758005938e-08,
      "loss": 0.3501,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7083
    },
    {
      "epoch": 0.9632852869186838,
      "grad_norm": 0.28125,
      "learning_rate": 8.144447786972986e-08,
      "loss": 0.5788,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7084
    },
    {
      "epoch": 0.9634212673375034,
      "grad_norm": 0.369140625,
      "learning_rate": 8.084088392308121e-08,
      "loss": 0.603,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7085
    },
    {
      "epoch": 0.9635572477563231,
      "grad_norm": 0.48828125,
      "learning_rate": 8.023952587616901e-08,
      "loss": 0.6683,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7086
    },
    {
      "epoch": 0.9636932281751428,
      "grad_norm": 0.357421875,
      "learning_rate": 7.964040386454818e-08,
      "loss": 0.703,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7087
    },
    {
      "epoch": 0.9638292085939625,
      "grad_norm": 0.62109375,
      "learning_rate": 7.904351802326626e-08,
      "loss": 0.9741,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7088
    },
    {
      "epoch": 0.9639651890127822,
      "grad_norm": 0.51171875,
      "learning_rate": 7.844886848687116e-08,
      "loss": 0.6733,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7089
    },
    {
      "epoch": 0.9641011694316018,
      "grad_norm": 0.40234375,
      "learning_rate": 7.785645538940234e-08,
      "loss": 0.7578,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7090
    },
    {
      "epoch": 0.9642371498504215,
      "grad_norm": 0.31640625,
      "learning_rate": 7.726627886439741e-08,
      "loss": 0.5441,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7091
    },
    {
      "epoch": 0.9643731302692412,
      "grad_norm": 0.546875,
      "learning_rate": 7.667833904488886e-08,
      "loss": 0.6862,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7092
    },
    {
      "epoch": 0.9645091106880609,
      "grad_norm": 0.453125,
      "learning_rate": 7.609263606340622e-08,
      "loss": 0.5804,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7093
    },
    {
      "epoch": 0.9646450911068806,
      "grad_norm": 0.375,
      "learning_rate": 7.550917005197389e-08,
      "loss": 0.7155,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7094
    },
    {
      "epoch": 0.9647810715257003,
      "grad_norm": 0.46484375,
      "learning_rate": 7.492794114211222e-08,
      "loss": 0.7228,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7095
    },
    {
      "epoch": 0.96491705194452,
      "grad_norm": 0.474609375,
      "learning_rate": 7.434894946483751e-08,
      "loss": 0.7469,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7096
    },
    {
      "epoch": 0.9650530323633397,
      "grad_norm": 0.64453125,
      "learning_rate": 7.377219515066092e-08,
      "loss": 0.5586,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7097
    },
    {
      "epoch": 0.9651890127821594,
      "grad_norm": 0.55859375,
      "learning_rate": 7.319767832959069e-08,
      "loss": 0.5807,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7098
    },
    {
      "epoch": 0.9653249932009791,
      "grad_norm": 0.330078125,
      "learning_rate": 7.262539913112987e-08,
      "loss": 0.6068,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7099
    },
    {
      "epoch": 0.9654609736197988,
      "grad_norm": 0.37109375,
      "learning_rate": 7.20553576842764e-08,
      "loss": 0.7135,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7100
    },
    {
      "epoch": 0.9655969540386184,
      "grad_norm": 0.3984375,
      "learning_rate": 7.14875541175264e-08,
      "loss": 0.6561,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7101
    },
    {
      "epoch": 0.9657329344574381,
      "grad_norm": 0.44921875,
      "learning_rate": 7.092198855886746e-08,
      "loss": 0.8646,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7102
    },
    {
      "epoch": 0.9658689148762578,
      "grad_norm": 0.318359375,
      "learning_rate": 7.035866113578649e-08,
      "loss": 0.486,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7103
    },
    {
      "epoch": 0.9660048952950775,
      "grad_norm": 0.4296875,
      "learning_rate": 6.979757197526305e-08,
      "loss": 0.688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7104
    },
    {
      "epoch": 0.9661408757138972,
      "grad_norm": 0.486328125,
      "learning_rate": 6.923872120377484e-08,
      "loss": 0.7069,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7105
    },
    {
      "epoch": 0.9662768561327169,
      "grad_norm": 0.376953125,
      "learning_rate": 6.868210894729333e-08,
      "loss": 0.5265,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7106
    },
    {
      "epoch": 0.9664128365515365,
      "grad_norm": 0.50390625,
      "learning_rate": 6.812773533128592e-08,
      "loss": 0.7492,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7107
    },
    {
      "epoch": 0.9665488169703562,
      "grad_norm": 0.392578125,
      "learning_rate": 6.757560048071376e-08,
      "loss": 0.5143,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7108
    },
    {
      "epoch": 0.966684797389176,
      "grad_norm": 0.55078125,
      "learning_rate": 6.70257045200362e-08,
      "loss": 0.7918,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7109
    },
    {
      "epoch": 0.9668207778079957,
      "grad_norm": 0.2890625,
      "learning_rate": 6.64780475732063e-08,
      "loss": 0.5181,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7110
    },
    {
      "epoch": 0.9669567582268154,
      "grad_norm": 0.55078125,
      "learning_rate": 6.593262976367309e-08,
      "loss": 0.6398,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7111
    },
    {
      "epoch": 0.967092738645635,
      "grad_norm": 0.2138671875,
      "learning_rate": 6.538945121437934e-08,
      "loss": 0.4111,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7112
    },
    {
      "epoch": 0.9672287190644547,
      "grad_norm": 0.447265625,
      "learning_rate": 6.484851204776488e-08,
      "loss": 0.6525,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7113
    },
    {
      "epoch": 0.9673646994832744,
      "grad_norm": 0.419921875,
      "learning_rate": 6.430981238576328e-08,
      "loss": 0.6687,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7114
    },
    {
      "epoch": 0.9675006799020941,
      "grad_norm": 1.078125,
      "learning_rate": 6.377335234980408e-08,
      "loss": 0.6344,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7115
    },
    {
      "epoch": 0.9676366603209138,
      "grad_norm": 0.392578125,
      "learning_rate": 6.323913206081277e-08,
      "loss": 0.6623,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7116
    },
    {
      "epoch": 0.9677726407397335,
      "grad_norm": 0.263671875,
      "learning_rate": 6.270715163920971e-08,
      "loss": 0.3328,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7117
    },
    {
      "epoch": 0.9679086211585531,
      "grad_norm": 0.390625,
      "learning_rate": 6.217741120490783e-08,
      "loss": 0.5638,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7118
    },
    {
      "epoch": 0.9680446015773728,
      "grad_norm": 0.375,
      "learning_rate": 6.164991087731831e-08,
      "loss": 0.5905,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7119
    },
    {
      "epoch": 0.9681805819961925,
      "grad_norm": 0.447265625,
      "learning_rate": 6.112465077534602e-08,
      "loss": 0.5905,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7120
    },
    {
      "epoch": 0.9683165624150122,
      "grad_norm": 0.466796875,
      "learning_rate": 6.06016310173907e-08,
      "loss": 0.7699,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7121
    },
    {
      "epoch": 0.968452542833832,
      "grad_norm": 0.369140625,
      "learning_rate": 6.008085172134804e-08,
      "loss": 0.659,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7122
    },
    {
      "epoch": 0.9685885232526517,
      "grad_norm": 0.46484375,
      "learning_rate": 5.956231300460747e-08,
      "loss": 0.6921,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7123
    },
    {
      "epoch": 0.9687245036714713,
      "grad_norm": 0.466796875,
      "learning_rate": 5.904601498405327e-08,
      "loss": 0.6045,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7124
    },
    {
      "epoch": 0.968860484090291,
      "grad_norm": 0.9765625,
      "learning_rate": 5.8531957776067905e-08,
      "loss": 0.6423,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7125
    },
    {
      "epoch": 0.9689964645091107,
      "grad_norm": 0.35546875,
      "learning_rate": 5.802014149652313e-08,
      "loss": 0.4189,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7126
    },
    {
      "epoch": 0.9691324449279304,
      "grad_norm": 0.28125,
      "learning_rate": 5.751056626078999e-08,
      "loss": 0.6416,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7127
    },
    {
      "epoch": 0.9692684253467501,
      "grad_norm": 0.76953125,
      "learning_rate": 5.700323218373327e-08,
      "loss": 0.7277,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7128
    },
    {
      "epoch": 0.9694044057655697,
      "grad_norm": 0.44140625,
      "learning_rate": 5.649813937971149e-08,
      "loss": 0.7427,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7129
    },
    {
      "epoch": 0.9695403861843894,
      "grad_norm": 0.482421875,
      "learning_rate": 5.5995287962579135e-08,
      "loss": 0.7188,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7130
    },
    {
      "epoch": 0.9696763666032091,
      "grad_norm": 0.58984375,
      "learning_rate": 5.5494678045685534e-08,
      "loss": 0.7376,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7131
    },
    {
      "epoch": 0.9698123470220288,
      "grad_norm": 0.451171875,
      "learning_rate": 5.4996309741873755e-08,
      "loss": 0.7375,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7132
    },
    {
      "epoch": 0.9699483274408485,
      "grad_norm": 0.423828125,
      "learning_rate": 5.4500183163481714e-08,
      "loss": 0.776,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7133
    },
    {
      "epoch": 0.9700843078596683,
      "grad_norm": 0.275390625,
      "learning_rate": 5.400629842234328e-08,
      "loss": 0.5075,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7134
    },
    {
      "epoch": 0.9702202882784879,
      "grad_norm": 0.419921875,
      "learning_rate": 5.351465562978608e-08,
      "loss": 0.592,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7135
    },
    {
      "epoch": 0.9703562686973076,
      "grad_norm": 0.28515625,
      "learning_rate": 5.302525489663146e-08,
      "loss": 0.6097,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7136
    },
    {
      "epoch": 0.9704922491161273,
      "grad_norm": 0.3515625,
      "learning_rate": 5.253809633319673e-08,
      "loss": 0.7227,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7137
    },
    {
      "epoch": 0.970628229534947,
      "grad_norm": 0.443359375,
      "learning_rate": 5.205318004929405e-08,
      "loss": 0.7233,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7138
    },
    {
      "epoch": 0.9707642099537667,
      "grad_norm": 0.376953125,
      "learning_rate": 5.15705061542282e-08,
      "loss": 0.672,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7139
    },
    {
      "epoch": 0.9709001903725863,
      "grad_norm": 0.28125,
      "learning_rate": 5.109007475680217e-08,
      "loss": 0.5823,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7140
    },
    {
      "epoch": 0.971036170791406,
      "grad_norm": 0.341796875,
      "learning_rate": 5.0611885965307086e-08,
      "loss": 0.5602,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7141
    },
    {
      "epoch": 0.9711721512102257,
      "grad_norm": 0.3828125,
      "learning_rate": 5.0135939887536736e-08,
      "loss": 0.73,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7142
    },
    {
      "epoch": 0.9713081316290454,
      "grad_norm": 0.349609375,
      "learning_rate": 4.9662236630771963e-08,
      "loss": 0.6063,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7143
    },
    {
      "epoch": 0.9714441120478651,
      "grad_norm": 0.65625,
      "learning_rate": 4.919077630179403e-08,
      "loss": 0.7606,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7144
    },
    {
      "epoch": 0.9715800924666848,
      "grad_norm": 0.30859375,
      "learning_rate": 4.8721559006873473e-08,
      "loss": 0.3643,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7145
    },
    {
      "epoch": 0.9717160728855044,
      "grad_norm": 0.53125,
      "learning_rate": 4.825458485178014e-08,
      "loss": 0.9007,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7146
    },
    {
      "epoch": 0.9718520533043242,
      "grad_norm": 0.2890625,
      "learning_rate": 4.778985394177204e-08,
      "loss": 0.5863,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7147
    },
    {
      "epoch": 0.9719880337231439,
      "grad_norm": 0.396484375,
      "learning_rate": 4.732736638160984e-08,
      "loss": 0.7362,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7148
    },
    {
      "epoch": 0.9721240141419636,
      "grad_norm": 0.2890625,
      "learning_rate": 4.686712227553902e-08,
      "loss": 0.6263,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7149
    },
    {
      "epoch": 0.9722599945607833,
      "grad_norm": 0.271484375,
      "learning_rate": 4.640912172730883e-08,
      "loss": 0.5648,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7150
    },
    {
      "epoch": 0.972395974979603,
      "grad_norm": 0.49609375,
      "learning_rate": 4.595336484015445e-08,
      "loss": 0.8118,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7151
    },
    {
      "epoch": 0.9725319553984226,
      "grad_norm": 0.62109375,
      "learning_rate": 4.5499851716811484e-08,
      "loss": 0.7702,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7152
    },
    {
      "epoch": 0.9726679358172423,
      "grad_norm": 0.291015625,
      "learning_rate": 4.504858245950594e-08,
      "loss": 0.5182,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7153
    },
    {
      "epoch": 0.972803916236062,
      "grad_norm": 0.294921875,
      "learning_rate": 4.4599557169959785e-08,
      "loss": 0.6087,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7154
    },
    {
      "epoch": 0.9729398966548817,
      "grad_norm": 0.349609375,
      "learning_rate": 4.4152775949388716e-08,
      "loss": 0.5207,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7155
    },
    {
      "epoch": 0.9730758770737014,
      "grad_norm": 0.50390625,
      "learning_rate": 4.370823889850439e-08,
      "loss": 0.7826,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7156
    },
    {
      "epoch": 0.973211857492521,
      "grad_norm": 0.65625,
      "learning_rate": 4.3265946117507783e-08,
      "loss": 0.4829,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7157
    },
    {
      "epoch": 0.9733478379113407,
      "grad_norm": 0.31640625,
      "learning_rate": 4.2825897706100237e-08,
      "loss": 0.596,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7158
    },
    {
      "epoch": 0.9734838183301605,
      "grad_norm": 0.5078125,
      "learning_rate": 4.23880937634713e-08,
      "loss": 0.8052,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7159
    },
    {
      "epoch": 0.9736197987489802,
      "grad_norm": 0.63671875,
      "learning_rate": 4.195253438830871e-08,
      "loss": 0.717,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7160
    },
    {
      "epoch": 0.9737557791677999,
      "grad_norm": 0.453125,
      "learning_rate": 4.15192196787928e-08,
      "loss": 0.4373,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7161
    },
    {
      "epoch": 0.9738917595866196,
      "grad_norm": 0.31640625,
      "learning_rate": 4.108814973259878e-08,
      "loss": 0.6247,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7162
    },
    {
      "epoch": 0.9740277400054392,
      "grad_norm": 0.74609375,
      "learning_rate": 4.065932464689337e-08,
      "loss": 0.5295,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7163
    },
    {
      "epoch": 0.9741637204242589,
      "grad_norm": 0.53515625,
      "learning_rate": 4.023274451834036e-08,
      "loss": 0.8397,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7164
    },
    {
      "epoch": 0.9742997008430786,
      "grad_norm": 0.54296875,
      "learning_rate": 3.9808409443095054e-08,
      "loss": 0.5467,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7165
    },
    {
      "epoch": 0.9744356812618983,
      "grad_norm": 0.1865234375,
      "learning_rate": 3.9386319516809826e-08,
      "loss": 0.3516,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7166
    },
    {
      "epoch": 0.974571661680718,
      "grad_norm": 0.4140625,
      "learning_rate": 3.896647483462634e-08,
      "loss": 0.6722,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7167
    },
    {
      "epoch": 0.9747076420995376,
      "grad_norm": 0.41796875,
      "learning_rate": 3.8548875491183355e-08,
      "loss": 0.7572,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7168
    },
    {
      "epoch": 0.9748436225183573,
      "grad_norm": 0.26953125,
      "learning_rate": 3.813352158061445e-08,
      "loss": 0.4861,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7169
    },
    {
      "epoch": 0.974979602937177,
      "grad_norm": 0.470703125,
      "learning_rate": 3.7720413196543625e-08,
      "loss": 0.6857,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7170
    },
    {
      "epoch": 0.9751155833559967,
      "grad_norm": 0.392578125,
      "learning_rate": 3.730955043209083e-08,
      "loss": 0.7303,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7171
    },
    {
      "epoch": 0.9752515637748165,
      "grad_norm": 0.267578125,
      "learning_rate": 3.6900933379870887e-08,
      "loss": 0.4318,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7172
    },
    {
      "epoch": 0.9753875441936362,
      "grad_norm": 0.267578125,
      "learning_rate": 3.6494562131990096e-08,
      "loss": 0.4836,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7173
    },
    {
      "epoch": 0.9755235246124558,
      "grad_norm": 0.4296875,
      "learning_rate": 3.609043678004964e-08,
      "loss": 0.7493,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7174
    },
    {
      "epoch": 0.9756595050312755,
      "grad_norm": 0.27734375,
      "learning_rate": 3.568855741514332e-08,
      "loss": 0.5103,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7175
    },
    {
      "epoch": 0.9757954854500952,
      "grad_norm": 0.279296875,
      "learning_rate": 3.5288924127862e-08,
      "loss": 0.3923,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7176
    },
    {
      "epoch": 0.9759314658689149,
      "grad_norm": 0.423828125,
      "learning_rate": 3.489153700828585e-08,
      "loss": 0.4967,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7177
    },
    {
      "epoch": 0.9760674462877346,
      "grad_norm": 0.515625,
      "learning_rate": 3.44963961459921e-08,
      "loss": 0.7318,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7178
    },
    {
      "epoch": 0.9762034267065542,
      "grad_norm": 0.48828125,
      "learning_rate": 3.410350163004839e-08,
      "loss": 0.8638,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7179
    },
    {
      "epoch": 0.9763394071253739,
      "grad_norm": 0.875,
      "learning_rate": 3.371285354902054e-08,
      "loss": 0.6755,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7180
    },
    {
      "epoch": 0.9764753875441936,
      "grad_norm": 0.310546875,
      "learning_rate": 3.332445199096368e-08,
      "loss": 0.5845,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7181
    },
    {
      "epoch": 0.9766113679630133,
      "grad_norm": 0.431640625,
      "learning_rate": 3.293829704342777e-08,
      "loss": 0.4958,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7182
    },
    {
      "epoch": 0.976747348381833,
      "grad_norm": 0.302734375,
      "learning_rate": 3.2554388793458735e-08,
      "loss": 0.5203,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7183
    },
    {
      "epoch": 0.9768833288006527,
      "grad_norm": 0.291015625,
      "learning_rate": 3.2172727327594024e-08,
      "loss": 0.515,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7184
    },
    {
      "epoch": 0.9770193092194724,
      "grad_norm": 0.375,
      "learning_rate": 3.179331273186259e-08,
      "loss": 0.6504,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7185
    },
    {
      "epoch": 0.9771552896382921,
      "grad_norm": 0.337890625,
      "learning_rate": 3.141614509179158e-08,
      "loss": 0.6403,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7186
    },
    {
      "epoch": 0.9772912700571118,
      "grad_norm": 0.28515625,
      "learning_rate": 3.1041224492397436e-08,
      "loss": 0.5887,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7187
    },
    {
      "epoch": 0.9774272504759315,
      "grad_norm": 0.7265625,
      "learning_rate": 3.066855101819255e-08,
      "loss": 0.639,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7188
    },
    {
      "epoch": 0.9775632308947512,
      "grad_norm": 0.337890625,
      "learning_rate": 3.029812475318306e-08,
      "loss": 0.6143,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7189
    },
    {
      "epoch": 0.9776992113135708,
      "grad_norm": 0.330078125,
      "learning_rate": 2.9929945780865496e-08,
      "loss": 0.5106,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7190
    },
    {
      "epoch": 0.9778351917323905,
      "grad_norm": 0.400390625,
      "learning_rate": 2.956401418423349e-08,
      "loss": 0.7538,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7191
    },
    {
      "epoch": 0.9779711721512102,
      "grad_norm": 0.349609375,
      "learning_rate": 2.9200330045772164e-08,
      "loss": 0.6087,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7192
    },
    {
      "epoch": 0.9781071525700299,
      "grad_norm": 0.279296875,
      "learning_rate": 2.883889344746038e-08,
      "loss": 0.5946,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7193
    },
    {
      "epoch": 0.9782431329888496,
      "grad_norm": 0.435546875,
      "learning_rate": 2.8479704470769642e-08,
      "loss": 0.6414,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7194
    },
    {
      "epoch": 0.9783791134076693,
      "grad_norm": 0.365234375,
      "learning_rate": 2.812276319666629e-08,
      "loss": 0.6935,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7195
    },
    {
      "epoch": 0.9785150938264889,
      "grad_norm": 0.326171875,
      "learning_rate": 2.7768069705609302e-08,
      "loss": 0.4113,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7196
    },
    {
      "epoch": 0.9786510742453087,
      "grad_norm": 0.306640625,
      "learning_rate": 2.7415624077551383e-08,
      "loss": 0.6611,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7197
    },
    {
      "epoch": 0.9787870546641284,
      "grad_norm": 0.349609375,
      "learning_rate": 2.7065426391935655e-08,
      "loss": 0.5719,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7198
    },
    {
      "epoch": 0.9789230350829481,
      "grad_norm": 0.392578125,
      "learning_rate": 2.6717476727704527e-08,
      "loss": 0.7575,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7199
    },
    {
      "epoch": 0.9790590155017678,
      "grad_norm": 0.42578125,
      "learning_rate": 2.6371775163287482e-08,
      "loss": 0.8467,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7200
    },
    {
      "epoch": 0.9791949959205875,
      "grad_norm": 0.345703125,
      "learning_rate": 2.6028321776609966e-08,
      "loss": 0.7337,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7201
    },
    {
      "epoch": 0.9793309763394071,
      "grad_norm": 0.408203125,
      "learning_rate": 2.568711664509227e-08,
      "loss": 0.7803,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7202
    },
    {
      "epoch": 0.9794669567582268,
      "grad_norm": 0.875,
      "learning_rate": 2.534815984564509e-08,
      "loss": 0.7726,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7203
    },
    {
      "epoch": 0.9796029371770465,
      "grad_norm": 0.267578125,
      "learning_rate": 2.5011451454672876e-08,
      "loss": 0.5085,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7204
    },
    {
      "epoch": 0.9797389175958662,
      "grad_norm": 0.35546875,
      "learning_rate": 2.4676991548074902e-08,
      "loss": 0.5074,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7205
    },
    {
      "epoch": 0.9798748980146859,
      "grad_norm": 0.421875,
      "learning_rate": 2.434478020124309e-08,
      "loss": 0.6707,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7206
    },
    {
      "epoch": 0.9800108784335055,
      "grad_norm": 0.38671875,
      "learning_rate": 2.4014817489059762e-08,
      "loss": 0.8781,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7207
    },
    {
      "epoch": 0.9801468588523252,
      "grad_norm": 0.38671875,
      "learning_rate": 2.3687103485904307e-08,
      "loss": 0.7793,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7208
    },
    {
      "epoch": 0.9802828392711449,
      "grad_norm": 0.298828125,
      "learning_rate": 2.3361638265646524e-08,
      "loss": 0.5145,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7209
    },
    {
      "epoch": 0.9804188196899647,
      "grad_norm": 0.40625,
      "learning_rate": 2.3038421901651064e-08,
      "loss": 0.6374,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7210
    },
    {
      "epoch": 0.9805548001087844,
      "grad_norm": 0.58203125,
      "learning_rate": 2.2717454466774092e-08,
      "loss": 0.6868,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7211
    },
    {
      "epoch": 0.9806907805276041,
      "grad_norm": 0.353515625,
      "learning_rate": 2.239873603336662e-08,
      "loss": 0.6318,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7212
    },
    {
      "epoch": 0.9808267609464237,
      "grad_norm": 1.2421875,
      "learning_rate": 2.2082266673270068e-08,
      "loss": 0.7624,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7213
    },
    {
      "epoch": 0.9809627413652434,
      "grad_norm": 0.361328125,
      "learning_rate": 2.1768046457821822e-08,
      "loss": 0.6346,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7214
    },
    {
      "epoch": 0.9810987217840631,
      "grad_norm": 0.466796875,
      "learning_rate": 2.1456075457850777e-08,
      "loss": 0.6826,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7215
    },
    {
      "epoch": 0.9812347022028828,
      "grad_norm": 0.404296875,
      "learning_rate": 2.114635374367846e-08,
      "loss": 0.7482,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7216
    },
    {
      "epoch": 0.9813706826217025,
      "grad_norm": 0.625,
      "learning_rate": 2.0838881385121245e-08,
      "loss": 0.555,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7217
    },
    {
      "epoch": 0.9815066630405221,
      "grad_norm": 0.330078125,
      "learning_rate": 2.0533658451484807e-08,
      "loss": 0.4993,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7218
    },
    {
      "epoch": 0.9816426434593418,
      "grad_norm": 0.2373046875,
      "learning_rate": 2.0230685011571883e-08,
      "loss": 0.4022,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7219
    },
    {
      "epoch": 0.9817786238781615,
      "grad_norm": 0.3125,
      "learning_rate": 1.9929961133676733e-08,
      "loss": 0.6049,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7220
    },
    {
      "epoch": 0.9819146042969812,
      "grad_norm": 0.318359375,
      "learning_rate": 1.9631486885585137e-08,
      "loss": 0.5999,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7221
    },
    {
      "epoch": 0.982050584715801,
      "grad_norm": 0.423828125,
      "learning_rate": 1.933526233457661e-08,
      "loss": 0.7974,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7222
    },
    {
      "epoch": 0.9821865651346207,
      "grad_norm": 0.390625,
      "learning_rate": 1.9041287547424404e-08,
      "loss": 0.5765,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7223
    },
    {
      "epoch": 0.9823225455534403,
      "grad_norm": 0.4296875,
      "learning_rate": 1.87495625903944e-08,
      "loss": 0.6774,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7224
    },
    {
      "epoch": 0.98245852597226,
      "grad_norm": 0.373046875,
      "learning_rate": 1.8460087529245107e-08,
      "loss": 0.6307,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7225
    },
    {
      "epoch": 0.9825945063910797,
      "grad_norm": 0.353515625,
      "learning_rate": 1.8172862429226557e-08,
      "loss": 0.6657,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7226
    },
    {
      "epoch": 0.9827304868098994,
      "grad_norm": 0.470703125,
      "learning_rate": 1.7887887355082513e-08,
      "loss": 0.6608,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7227
    },
    {
      "epoch": 0.9828664672287191,
      "grad_norm": 0.498046875,
      "learning_rate": 1.7605162371051587e-08,
      "loss": 0.8581,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7228
    },
    {
      "epoch": 0.9830024476475387,
      "grad_norm": 0.380859375,
      "learning_rate": 1.7324687540862804e-08,
      "loss": 0.6336,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7229
    },
    {
      "epoch": 0.9831384280663584,
      "grad_norm": 0.59765625,
      "learning_rate": 1.7046462927738926e-08,
      "loss": 0.6688,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7230
    },
    {
      "epoch": 0.9832744084851781,
      "grad_norm": 0.361328125,
      "learning_rate": 1.677048859439423e-08,
      "loss": 0.6304,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7231
    },
    {
      "epoch": 0.9834103889039978,
      "grad_norm": 0.384765625,
      "learning_rate": 1.649676460303673e-08,
      "loss": 0.6338,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7232
    },
    {
      "epoch": 0.9835463693228175,
      "grad_norm": 0.69921875,
      "learning_rate": 1.6225291015368183e-08,
      "loss": 0.8159,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7233
    },
    {
      "epoch": 0.9836823497416372,
      "grad_norm": 0.349609375,
      "learning_rate": 1.595606789258075e-08,
      "loss": 0.5335,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7234
    },
    {
      "epoch": 0.9838183301604569,
      "grad_norm": 0.50390625,
      "learning_rate": 1.5689095295361444e-08,
      "loss": 0.7357,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7235
    },
    {
      "epoch": 0.9839543105792766,
      "grad_norm": 0.26953125,
      "learning_rate": 1.5424373283889904e-08,
      "loss": 0.4556,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7236
    },
    {
      "epoch": 0.9840902909980963,
      "grad_norm": 0.228515625,
      "learning_rate": 1.5161901917836175e-08,
      "loss": 0.4836,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7237
    },
    {
      "epoch": 0.984226271416916,
      "grad_norm": 0.4296875,
      "learning_rate": 1.4901681256365153e-08,
      "loss": 0.6745,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7238
    },
    {
      "epoch": 0.9843622518357357,
      "grad_norm": 0.3125,
      "learning_rate": 1.464371135813436e-08,
      "loss": 0.6045,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7239
    },
    {
      "epoch": 0.9844982322545553,
      "grad_norm": 0.296875,
      "learning_rate": 1.438799228129173e-08,
      "loss": 0.6138,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7240
    },
    {
      "epoch": 0.984634212673375,
      "grad_norm": 0.486328125,
      "learning_rate": 1.413452408348115e-08,
      "loss": 0.6991,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7241
    },
    {
      "epoch": 0.9847701930921947,
      "grad_norm": 0.296875,
      "learning_rate": 1.3883306821836918e-08,
      "loss": 0.4803,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7242
    },
    {
      "epoch": 0.9849061735110144,
      "grad_norm": 0.29296875,
      "learning_rate": 1.3634340552985958e-08,
      "loss": 0.6506,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7243
    },
    {
      "epoch": 0.9850421539298341,
      "grad_norm": 0.490234375,
      "learning_rate": 1.3387625333048936e-08,
      "loss": 0.5083,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7244
    },
    {
      "epoch": 0.9851781343486538,
      "grad_norm": 0.322265625,
      "learning_rate": 1.3143161217639144e-08,
      "loss": 0.6115,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7245
    },
    {
      "epoch": 0.9853141147674734,
      "grad_norm": 0.443359375,
      "learning_rate": 1.290094826186028e-08,
      "loss": 0.6357,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7246
    },
    {
      "epoch": 0.9854500951862932,
      "grad_norm": 0.546875,
      "learning_rate": 1.2660986520309781e-08,
      "loss": 0.7983,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7247
    },
    {
      "epoch": 0.9855860756051129,
      "grad_norm": 0.39453125,
      "learning_rate": 1.2423276047079936e-08,
      "loss": 0.7541,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7248
    },
    {
      "epoch": 0.9857220560239326,
      "grad_norm": 0.359375,
      "learning_rate": 1.2187816895752324e-08,
      "loss": 0.4966,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7249
    },
    {
      "epoch": 0.9858580364427523,
      "grad_norm": 0.41796875,
      "learning_rate": 1.1954609119403382e-08,
      "loss": 0.759,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7250
    },
    {
      "epoch": 0.985994016861572,
      "grad_norm": 0.263671875,
      "learning_rate": 1.1723652770598836e-08,
      "loss": 0.5094,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7251
    },
    {
      "epoch": 0.9861299972803916,
      "grad_norm": 0.322265625,
      "learning_rate": 1.1494947901401487e-08,
      "loss": 0.6128,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7252
    },
    {
      "epoch": 0.9862659776992113,
      "grad_norm": 0.56640625,
      "learning_rate": 1.126849456336232e-08,
      "loss": 0.8371,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7253
    },
    {
      "epoch": 0.986401958118031,
      "grad_norm": 0.3125,
      "learning_rate": 1.1044292807528278e-08,
      "loss": 0.6457,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7254
    },
    {
      "epoch": 0.9865379385368507,
      "grad_norm": 0.365234375,
      "learning_rate": 1.0822342684435605e-08,
      "loss": 0.7723,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7255
    },
    {
      "epoch": 0.9866739189556704,
      "grad_norm": 0.322265625,
      "learning_rate": 1.06026442441165e-08,
      "loss": 0.605,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7256
    },
    {
      "epoch": 0.98680989937449,
      "grad_norm": 0.310546875,
      "learning_rate": 1.0385197536091351e-08,
      "loss": 0.585,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7257
    },
    {
      "epoch": 0.9869458797933097,
      "grad_norm": 0.37890625,
      "learning_rate": 1.0170002609377616e-08,
      "loss": 0.6577,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7258
    },
    {
      "epoch": 0.9870818602121294,
      "grad_norm": 0.3671875,
      "learning_rate": 9.957059512480938e-09,
      "loss": 0.6517,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7259
    },
    {
      "epoch": 0.9872178406309492,
      "grad_norm": 0.4296875,
      "learning_rate": 9.746368293401809e-09,
      "loss": 0.7051,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7260
    },
    {
      "epoch": 0.9873538210497689,
      "grad_norm": 0.5546875,
      "learning_rate": 9.53792899963335e-09,
      "loss": 0.661,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7261
    },
    {
      "epoch": 0.9874898014685886,
      "grad_norm": 0.80078125,
      "learning_rate": 9.3317416781602e-09,
      "loss": 0.7466,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7262
    },
    {
      "epoch": 0.9876257818874082,
      "grad_norm": 0.328125,
      "learning_rate": 9.127806375458514e-09,
      "loss": 0.5833,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7263
    },
    {
      "epoch": 0.9877617623062279,
      "grad_norm": 0.3515625,
      "learning_rate": 8.926123137499298e-09,
      "loss": 0.6353,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7264
    },
    {
      "epoch": 0.9878977427250476,
      "grad_norm": 0.5,
      "learning_rate": 8.726692009742855e-09,
      "loss": 0.66,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7265
    },
    {
      "epoch": 0.9880337231438673,
      "grad_norm": 0.486328125,
      "learning_rate": 8.529513037144333e-09,
      "loss": 0.5181,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7266
    },
    {
      "epoch": 0.988169703562687,
      "grad_norm": 0.51953125,
      "learning_rate": 8.334586264150401e-09,
      "loss": 0.7362,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7267
    },
    {
      "epoch": 0.9883056839815066,
      "grad_norm": 0.388671875,
      "learning_rate": 8.141911734699248e-09,
      "loss": 0.647,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7268
    },
    {
      "epoch": 0.9884416644003263,
      "grad_norm": 0.369140625,
      "learning_rate": 7.951489492222797e-09,
      "loss": 0.6719,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7269
    },
    {
      "epoch": 0.988577644819146,
      "grad_norm": 0.185546875,
      "learning_rate": 7.76331957964449e-09,
      "loss": 0.3774,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7270
    },
    {
      "epoch": 0.9887136252379657,
      "grad_norm": 0.470703125,
      "learning_rate": 7.57740203937929e-09,
      "loss": 0.5378,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7271
    },
    {
      "epoch": 0.9888496056567854,
      "grad_norm": 0.34765625,
      "learning_rate": 7.3937369133358914e-09,
      "loss": 0.6213,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7272
    },
    {
      "epoch": 0.9889855860756052,
      "grad_norm": 0.26953125,
      "learning_rate": 7.212324242914514e-09,
      "loss": 0.5042,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7273
    },
    {
      "epoch": 0.9891215664944248,
      "grad_norm": 0.27734375,
      "learning_rate": 7.033164069008003e-09,
      "loss": 0.4004,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7274
    },
    {
      "epoch": 0.9892575469132445,
      "grad_norm": 0.52734375,
      "learning_rate": 6.856256432000719e-09,
      "loss": 0.4466,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7275
    },
    {
      "epoch": 0.9893935273320642,
      "grad_norm": 0.435546875,
      "learning_rate": 6.681601371770763e-09,
      "loss": 0.7614,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7276
    },
    {
      "epoch": 0.9895295077508839,
      "grad_norm": 0.3984375,
      "learning_rate": 6.5091989276866445e-09,
      "loss": 0.5947,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7277
    },
    {
      "epoch": 0.9896654881697036,
      "grad_norm": 0.41015625,
      "learning_rate": 6.339049138610609e-09,
      "loss": 0.6299,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7278
    },
    {
      "epoch": 0.9898014685885232,
      "grad_norm": 0.6953125,
      "learning_rate": 6.171152042895312e-09,
      "loss": 0.6017,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7279
    },
    {
      "epoch": 0.9899374490073429,
      "grad_norm": 0.50390625,
      "learning_rate": 6.005507678388256e-09,
      "loss": 0.9302,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7280
    },
    {
      "epoch": 0.9900734294261626,
      "grad_norm": 0.380859375,
      "learning_rate": 5.842116082427351e-09,
      "loss": 0.7192,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7281
    },
    {
      "epoch": 0.9902094098449823,
      "grad_norm": 0.296875,
      "learning_rate": 5.680977291843137e-09,
      "loss": 0.5648,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7282
    },
    {
      "epoch": 0.990345390263802,
      "grad_norm": 0.6015625,
      "learning_rate": 5.5220913429576696e-09,
      "loss": 0.8332,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7283
    },
    {
      "epoch": 0.9904813706826217,
      "grad_norm": 0.275390625,
      "learning_rate": 5.365458271585633e-09,
      "loss": 0.516,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7284
    },
    {
      "epoch": 0.9906173511014414,
      "grad_norm": 0.38671875,
      "learning_rate": 5.21107811303545e-09,
      "loss": 0.6619,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7285
    },
    {
      "epoch": 0.9907533315202611,
      "grad_norm": 0.412109375,
      "learning_rate": 5.058950902104842e-09,
      "loss": 0.605,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7286
    },
    {
      "epoch": 0.9908893119390808,
      "grad_norm": 0.43359375,
      "learning_rate": 4.909076673085267e-09,
      "loss": 0.8289,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7287
    },
    {
      "epoch": 0.9910252923579005,
      "grad_norm": 0.404296875,
      "learning_rate": 4.761455459760811e-09,
      "loss": 0.7201,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7288
    },
    {
      "epoch": 0.9911612727767202,
      "grad_norm": 0.333984375,
      "learning_rate": 4.616087295407079e-09,
      "loss": 0.6016,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7289
    },
    {
      "epoch": 0.9912972531955399,
      "grad_norm": 0.73046875,
      "learning_rate": 4.4729722127911934e-09,
      "loss": 0.6611,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7290
    },
    {
      "epoch": 0.9914332336143595,
      "grad_norm": 5.46875,
      "learning_rate": 4.332110244174015e-09,
      "loss": 0.6313,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7291
    },
    {
      "epoch": 0.9915692140331792,
      "grad_norm": 0.2412109375,
      "learning_rate": 4.193501421306811e-09,
      "loss": 0.3556,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7292
    },
    {
      "epoch": 0.9917051944519989,
      "grad_norm": 0.490234375,
      "learning_rate": 4.05714577543459e-09,
      "loss": 0.6221,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7293
    },
    {
      "epoch": 0.9918411748708186,
      "grad_norm": 0.2216796875,
      "learning_rate": 3.923043337291654e-09,
      "loss": 0.4414,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7294
    },
    {
      "epoch": 0.9919771552896383,
      "grad_norm": 0.337890625,
      "learning_rate": 3.7911941371082674e-09,
      "loss": 0.5516,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7295
    },
    {
      "epoch": 0.9921131357084579,
      "grad_norm": 0.498046875,
      "learning_rate": 3.661598204603989e-09,
      "loss": 0.6799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7296
    },
    {
      "epoch": 0.9922491161272776,
      "grad_norm": 0.53515625,
      "learning_rate": 3.5342555689921177e-09,
      "loss": 0.9676,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7297
    },
    {
      "epoch": 0.9923850965460974,
      "grad_norm": 0.443359375,
      "learning_rate": 3.4091662589752494e-09,
      "loss": 0.665,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7298
    },
    {
      "epoch": 0.9925210769649171,
      "grad_norm": 0.76953125,
      "learning_rate": 3.2863303027530492e-09,
      "loss": 0.8717,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7299
    },
    {
      "epoch": 0.9926570573837368,
      "grad_norm": 0.375,
      "learning_rate": 3.165747728011148e-09,
      "loss": 0.6016,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7300
    },
    {
      "epoch": 0.9927930378025565,
      "grad_norm": 0.47265625,
      "learning_rate": 3.0474185619333573e-09,
      "loss": 0.7399,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7301
    },
    {
      "epoch": 0.9929290182213761,
      "grad_norm": 0.66015625,
      "learning_rate": 2.9313428311894542e-09,
      "loss": 0.6981,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7302
    },
    {
      "epoch": 0.9930649986401958,
      "grad_norm": 0.302734375,
      "learning_rate": 2.8175205619462853e-09,
      "loss": 0.4207,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7303
    },
    {
      "epoch": 0.9932009790590155,
      "grad_norm": 0.41796875,
      "learning_rate": 2.7059517798611046e-09,
      "loss": 0.71,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7304
    },
    {
      "epoch": 0.9933369594778352,
      "grad_norm": 0.369140625,
      "learning_rate": 2.596636510081574e-09,
      "loss": 0.6698,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7305
    },
    {
      "epoch": 0.9934729398966549,
      "grad_norm": 0.58203125,
      "learning_rate": 2.489574777249093e-09,
      "loss": 0.7716,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7306
    },
    {
      "epoch": 0.9936089203154745,
      "grad_norm": 0.294921875,
      "learning_rate": 2.38476660549658e-09,
      "loss": 0.5998,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7307
    },
    {
      "epoch": 0.9937449007342942,
      "grad_norm": 0.296875,
      "learning_rate": 2.282212018449581e-09,
      "loss": 0.5488,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7308
    },
    {
      "epoch": 0.9938808811531139,
      "grad_norm": 0.7734375,
      "learning_rate": 2.1819110392251597e-09,
      "loss": 0.9181,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7309
    },
    {
      "epoch": 0.9940168615719337,
      "grad_norm": 0.2412109375,
      "learning_rate": 2.0838636904318974e-09,
      "loss": 0.3257,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7310
    },
    {
      "epoch": 0.9941528419907534,
      "grad_norm": 0.400390625,
      "learning_rate": 1.9880699941710048e-09,
      "loss": 0.8262,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7311
    },
    {
      "epoch": 0.9942888224095731,
      "grad_norm": 0.287109375,
      "learning_rate": 1.8945299720352086e-09,
      "loss": 0.6011,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7312
    },
    {
      "epoch": 0.9944248028283927,
      "grad_norm": 0.4921875,
      "learning_rate": 1.8032436451109746e-09,
      "loss": 0.8154,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7313
    },
    {
      "epoch": 0.9945607832472124,
      "grad_norm": 0.91796875,
      "learning_rate": 1.7142110339740669e-09,
      "loss": 0.8638,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7314
    },
    {
      "epoch": 0.9946967636660321,
      "grad_norm": 0.322265625,
      "learning_rate": 1.6274321586928765e-09,
      "loss": 0.6785,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7315
    },
    {
      "epoch": 0.9948327440848518,
      "grad_norm": 0.373046875,
      "learning_rate": 1.5429070388306434e-09,
      "loss": 0.7204,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7316
    },
    {
      "epoch": 0.9949687245036715,
      "grad_norm": 0.5,
      "learning_rate": 1.4606356934387944e-09,
      "loss": 0.5552,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7317
    },
    {
      "epoch": 0.9951047049224911,
      "grad_norm": 0.27734375,
      "learning_rate": 1.3806181410624952e-09,
      "loss": 0.5934,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7318
    },
    {
      "epoch": 0.9952406853413108,
      "grad_norm": 0.349609375,
      "learning_rate": 1.302854399738429e-09,
      "loss": 0.5238,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7319
    },
    {
      "epoch": 0.9953766657601305,
      "grad_norm": 0.2216796875,
      "learning_rate": 1.2273444869959072e-09,
      "loss": 0.4198,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7320
    },
    {
      "epoch": 0.9955126461789502,
      "grad_norm": 0.5,
      "learning_rate": 1.154088419855759e-09,
      "loss": 0.8483,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7321
    },
    {
      "epoch": 0.9956486265977699,
      "grad_norm": 0.478515625,
      "learning_rate": 1.0830862148314413e-09,
      "loss": 0.8921,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7322
    },
    {
      "epoch": 0.9957846070165897,
      "grad_norm": 0.439453125,
      "learning_rate": 1.0143378879257093e-09,
      "loss": 0.7799,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7323
    },
    {
      "epoch": 0.9959205874354093,
      "grad_norm": 0.484375,
      "learning_rate": 9.47843454637276e-10,
      "loss": 0.6916,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7324
    },
    {
      "epoch": 0.996056567854229,
      "grad_norm": 0.4453125,
      "learning_rate": 8.836029299552628e-10,
      "loss": 0.675,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7325
    },
    {
      "epoch": 0.9961925482730487,
      "grad_norm": 0.9296875,
      "learning_rate": 8.21616328358088e-10,
      "loss": 0.5949,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7326
    },
    {
      "epoch": 0.9963285286918684,
      "grad_norm": 0.3125,
      "learning_rate": 7.618836638190186e-10,
      "loss": 0.638,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7327
    },
    {
      "epoch": 0.9964645091106881,
      "grad_norm": 0.3125,
      "learning_rate": 7.044049498028393e-10,
      "loss": 0.6576,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7328
    },
    {
      "epoch": 0.9966004895295077,
      "grad_norm": 0.447265625,
      "learning_rate": 6.491801992669633e-10,
      "loss": 0.8953,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7329
    },
    {
      "epoch": 0.9967364699483274,
      "grad_norm": 0.251953125,
      "learning_rate": 5.962094246581007e-10,
      "loss": 0.5161,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7330
    },
    {
      "epoch": 0.9968724503671471,
      "grad_norm": 0.75390625,
      "learning_rate": 5.454926379167003e-10,
      "loss": 0.6047,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7331
    },
    {
      "epoch": 0.9970084307859668,
      "grad_norm": 0.423828125,
      "learning_rate": 4.970298504758386e-10,
      "loss": 0.741,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7332
    },
    {
      "epoch": 0.9971444112047865,
      "grad_norm": 0.62890625,
      "learning_rate": 4.508210732601104e-10,
      "loss": 0.5462,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7333
    },
    {
      "epoch": 0.9972803916236062,
      "grad_norm": 0.53125,
      "learning_rate": 4.068663166834075e-10,
      "loss": 0.783,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7334
    },
    {
      "epoch": 0.9974163720424258,
      "grad_norm": 0.26953125,
      "learning_rate": 3.6516559065558064e-10,
      "loss": 0.533,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7335
    },
    {
      "epoch": 0.9975523524612456,
      "grad_norm": 0.5859375,
      "learning_rate": 3.2571890457466783e-10,
      "loss": 0.6906,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7336
    },
    {
      "epoch": 0.9976883328800653,
      "grad_norm": 0.47265625,
      "learning_rate": 2.8852626733466557e-10,
      "loss": 0.702,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7337
    },
    {
      "epoch": 0.997824313298885,
      "grad_norm": 0.61328125,
      "learning_rate": 2.5358768731775783e-10,
      "loss": 0.66,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7338
    },
    {
      "epoch": 0.9979602937177047,
      "grad_norm": 0.400390625,
      "learning_rate": 2.2090317239875648e-10,
      "loss": 0.7363,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7339
    },
    {
      "epoch": 0.9980962741365244,
      "grad_norm": 0.330078125,
      "learning_rate": 1.904727299473219e-10,
      "loss": 0.6437,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7340
    },
    {
      "epoch": 0.998232254555344,
      "grad_norm": 0.279296875,
      "learning_rate": 1.62296366822412e-10,
      "loss": 0.4985,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7341
    },
    {
      "epoch": 0.9983682349741637,
      "grad_norm": 0.75390625,
      "learning_rate": 1.3637408937339224e-10,
      "loss": 0.5636,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7342
    },
    {
      "epoch": 0.9985042153929834,
      "grad_norm": 0.228515625,
      "learning_rate": 1.1270590344558685e-10,
      "loss": 0.3901,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7343
    },
    {
      "epoch": 0.9986401958118031,
      "grad_norm": 0.435546875,
      "learning_rate": 9.129181437361745e-11,
      "loss": 0.7904,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7344
    },
    {
      "epoch": 0.9987761762306228,
      "grad_norm": 0.36328125,
      "learning_rate": 7.213182698362353e-11,
      "loss": 0.769,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7345
    },
    {
      "epoch": 0.9989121566494424,
      "grad_norm": 0.240234375,
      "learning_rate": 5.522594559548289e-11,
      "loss": 0.5444,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7346
    },
    {
      "epoch": 0.9990481370682621,
      "grad_norm": 0.85546875,
      "learning_rate": 4.057417401837071e-11,
      "loss": 0.7067,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7347
    },
    {
      "epoch": 0.9991841174870819,
      "grad_norm": 0.40625,
      "learning_rate": 2.8176515557420957e-11,
      "loss": 0.7075,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7348
    },
    {
      "epoch": 0.9993200979059016,
      "grad_norm": 0.443359375,
      "learning_rate": 1.8032973004844524e-11,
      "loss": 0.7359,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7349
    },
    {
      "epoch": 0.9994560783247213,
      "grad_norm": 0.28515625,
      "learning_rate": 1.0143548648811064e-11,
      "loss": 0.5378,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7350
    },
    {
      "epoch": 0.999592058743541,
      "grad_norm": 0.453125,
      "learning_rate": 4.5082442678978655e-12,
      "loss": 0.6157,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7351
    },
    {
      "epoch": 0.9997280391623606,
      "grad_norm": 0.625,
      "learning_rate": 1.127061129979623e-12,
      "loss": 0.6589,
      "memory/device_mem_reserved(gib)": 47.75,
      "memory/max_mem_active(gib)": 39.5,
      "memory/max_mem_allocated(gib)": 39.5,
      "step": 7352
    }
  ],
  "logging_steps": 1,
  "max_steps": 7352,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.5772130034993217e+21,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}