{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0025188916876575,
  "eval_steps": 500,
  "global_step": 398,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0025188916876574307,
      "grad_norm": 29.875,
      "learning_rate": 2.3076923076923078e-07,
      "loss": 1.1914,
      "memory/device_mem_reserved(gib)": 80.65,
      "memory/max_mem_active(gib)": 67.28,
      "memory/max_mem_allocated(gib)": 67.28,
      "step": 1
    },
    {
      "epoch": 0.005037783375314861,
      "grad_norm": 26.875,
      "learning_rate": 4.6153846153846156e-07,
      "loss": 1.1719,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 2
    },
    {
      "epoch": 0.007556675062972292,
      "grad_norm": 28.0,
      "learning_rate": 6.923076923076923e-07,
      "loss": 1.166,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 3
    },
    {
      "epoch": 0.010075566750629723,
      "grad_norm": 61.25,
      "learning_rate": 9.230769230769231e-07,
      "loss": 1.6523,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 4
    },
    {
      "epoch": 0.012594458438287154,
      "grad_norm": 7.1875,
      "learning_rate": 1.1538461538461538e-06,
      "loss": 1.1035,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 5
    },
    {
      "epoch": 0.015113350125944584,
      "grad_norm": 6.0625,
      "learning_rate": 1.3846153846153846e-06,
      "loss": 1.1689,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 6
    },
    {
      "epoch": 0.017632241813602016,
      "grad_norm": 5.03125,
      "learning_rate": 1.6153846153846154e-06,
      "loss": 1.0928,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 7
    },
    {
      "epoch": 0.020151133501259445,
      "grad_norm": 9.3125,
      "learning_rate": 1.8461538461538462e-06,
      "loss": 1.0908,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 8
    },
    {
      "epoch": 0.022670025188916875,
      "grad_norm": 4.625,
      "learning_rate": 2.076923076923077e-06,
      "loss": 1.0566,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 9
    },
    {
      "epoch": 0.02518891687657431,
      "grad_norm": 3.59375,
      "learning_rate": 2.3076923076923077e-06,
      "loss": 1.0049,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 10
    },
    {
      "epoch": 0.027707808564231738,
      "grad_norm": 2.40625,
      "learning_rate": 2.5384615384615385e-06,
      "loss": 1.0156,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 11
    },
    {
      "epoch": 0.030226700251889168,
      "grad_norm": 2.34375,
      "learning_rate": 2.7692307692307693e-06,
      "loss": 0.9209,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 12
    },
    {
      "epoch": 0.0327455919395466,
      "grad_norm": 4.71875,
      "learning_rate": 3e-06,
      "loss": 1.1035,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 13
    },
    {
      "epoch": 0.03526448362720403,
      "grad_norm": 3.09375,
      "learning_rate": 3.230769230769231e-06,
      "loss": 1.0547,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 14
    },
    {
      "epoch": 0.037783375314861464,
      "grad_norm": 1.765625,
      "learning_rate": 3.4615384615384617e-06,
      "loss": 0.9268,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 15
    },
    {
      "epoch": 0.04030226700251889,
      "grad_norm": 2.171875,
      "learning_rate": 3.6923076923076925e-06,
      "loss": 1.0215,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 16
    },
    {
      "epoch": 0.042821158690176324,
      "grad_norm": 5.1875,
      "learning_rate": 3.923076923076923e-06,
      "loss": 0.9707,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 17
    },
    {
      "epoch": 0.04534005037783375,
      "grad_norm": 4.09375,
      "learning_rate": 4.153846153846154e-06,
      "loss": 0.959,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 18
    },
    {
      "epoch": 0.04785894206549118,
      "grad_norm": 3.421875,
      "learning_rate": 4.384615384615385e-06,
      "loss": 1.0137,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 19
    },
    {
      "epoch": 0.05037783375314862,
      "grad_norm": 2.71875,
      "learning_rate": 4.615384615384615e-06,
      "loss": 1.0137,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 20
    },
    {
      "epoch": 0.05289672544080604,
      "grad_norm": 1.6796875,
      "learning_rate": 4.8461538461538465e-06,
      "loss": 0.9062,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 21
    },
    {
      "epoch": 0.055415617128463476,
      "grad_norm": 2.109375,
      "learning_rate": 5.076923076923077e-06,
      "loss": 0.875,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 22
    },
    {
      "epoch": 0.05793450881612091,
      "grad_norm": 2.609375,
      "learning_rate": 5.307692307692308e-06,
      "loss": 0.9893,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 23
    },
    {
      "epoch": 0.060453400503778336,
      "grad_norm": 5.71875,
      "learning_rate": 5.5384615384615385e-06,
      "loss": 1.0166,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 24
    },
    {
      "epoch": 0.06297229219143577,
      "grad_norm": 1.640625,
      "learning_rate": 5.769230769230769e-06,
      "loss": 0.9834,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 25
    },
    {
      "epoch": 0.0654911838790932,
      "grad_norm": 1.390625,
      "learning_rate": 6e-06,
      "loss": 0.8379,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 26
    },
    {
      "epoch": 0.06801007556675064,
      "grad_norm": 1.4765625,
      "learning_rate": 6.2307692307692305e-06,
      "loss": 0.9717,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 27
    },
    {
      "epoch": 0.07052896725440806,
      "grad_norm": 1.453125,
      "learning_rate": 6.461538461538462e-06,
      "loss": 0.873,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 28
    },
    {
      "epoch": 0.07304785894206549,
      "grad_norm": 1.40625,
      "learning_rate": 6.692307692307692e-06,
      "loss": 0.9424,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 29
    },
    {
      "epoch": 0.07556675062972293,
      "grad_norm": 2.21875,
      "learning_rate": 6.923076923076923e-06,
      "loss": 0.9629,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 30
    },
    {
      "epoch": 0.07808564231738035,
      "grad_norm": 1.6875,
      "learning_rate": 7.153846153846154e-06,
      "loss": 0.8477,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 31
    },
    {
      "epoch": 0.08060453400503778,
      "grad_norm": 1.390625,
      "learning_rate": 7.384615384615385e-06,
      "loss": 0.9854,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 32
    },
    {
      "epoch": 0.08312342569269521,
      "grad_norm": 1.609375,
      "learning_rate": 7.615384615384615e-06,
      "loss": 0.9248,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 33
    },
    {
      "epoch": 0.08564231738035265,
      "grad_norm": 1.3828125,
      "learning_rate": 7.846153846153847e-06,
      "loss": 0.9873,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 34
    },
    {
      "epoch": 0.08816120906801007,
      "grad_norm": 1.2890625,
      "learning_rate": 8.076923076923077e-06,
      "loss": 0.8662,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 35
    },
    {
      "epoch": 0.0906801007556675,
      "grad_norm": 1.984375,
      "learning_rate": 8.307692307692307e-06,
      "loss": 0.835,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 36
    },
    {
      "epoch": 0.09319899244332494,
      "grad_norm": 1.546875,
      "learning_rate": 8.53846153846154e-06,
      "loss": 0.9141,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 37
    },
    {
      "epoch": 0.09571788413098237,
      "grad_norm": 1.453125,
      "learning_rate": 8.76923076923077e-06,
      "loss": 0.9336,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 38
    },
    {
      "epoch": 0.0982367758186398,
      "grad_norm": 1.3515625,
      "learning_rate": 9e-06,
      "loss": 0.959,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 39
    },
    {
      "epoch": 0.10075566750629723,
      "grad_norm": 1.3828125,
      "learning_rate": 8.998925871900279e-06,
      "loss": 0.8184,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 40
    },
    {
      "epoch": 0.10327455919395466,
      "grad_norm": 1.4765625,
      "learning_rate": 8.997849176845461e-06,
      "loss": 0.9004,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 41
    },
    {
      "epoch": 0.10579345088161209,
      "grad_norm": 2.1875,
      "learning_rate": 8.996769905622757e-06,
      "loss": 0.9365,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 42
    },
    {
      "epoch": 0.10831234256926953,
      "grad_norm": 1.6953125,
      "learning_rate": 8.995688048975247e-06,
      "loss": 0.9414,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 43
    },
    {
      "epoch": 0.11083123425692695,
      "grad_norm": 2.234375,
      "learning_rate": 8.994603597601599e-06,
      "loss": 1.0234,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 44
    },
    {
      "epoch": 0.11335012594458438,
      "grad_norm": 1.390625,
      "learning_rate": 8.993516542155818e-06,
      "loss": 0.8896,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 45
    },
    {
      "epoch": 0.11586901763224182,
      "grad_norm": 1.1328125,
      "learning_rate": 8.992426873246962e-06,
      "loss": 0.8965,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 46
    },
    {
      "epoch": 0.11838790931989925,
      "grad_norm": 1.703125,
      "learning_rate": 8.991334581438888e-06,
      "loss": 0.8301,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 47
    },
    {
      "epoch": 0.12090680100755667,
      "grad_norm": 1.875,
      "learning_rate": 8.990239657249966e-06,
      "loss": 0.8691,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 48
    },
    {
      "epoch": 0.12342569269521411,
      "grad_norm": 1.4375,
      "learning_rate": 8.989142091152815e-06,
      "loss": 0.9229,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 49
    },
    {
      "epoch": 0.12594458438287154,
      "grad_norm": 1.390625,
      "learning_rate": 8.988041873574018e-06,
      "loss": 0.9941,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 50
    },
    {
      "epoch": 0.12846347607052896,
      "grad_norm": 1.5625,
      "learning_rate": 8.986938994893847e-06,
      "loss": 0.8838,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 51
    },
    {
      "epoch": 0.1309823677581864,
      "grad_norm": 1.234375,
      "learning_rate": 8.985833445445984e-06,
      "loss": 0.8643,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 52
    },
    {
      "epoch": 0.13350125944584382,
      "grad_norm": 1.515625,
      "learning_rate": 8.984725215517241e-06,
      "loss": 0.9756,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 53
    },
    {
      "epoch": 0.13602015113350127,
      "grad_norm": 1.453125,
      "learning_rate": 8.98361429534727e-06,
      "loss": 0.8311,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 54
    },
    {
      "epoch": 0.1385390428211587,
      "grad_norm": 1.46875,
      "learning_rate": 8.982500675128276e-06,
      "loss": 0.8545,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 55
    },
    {
      "epoch": 0.14105793450881612,
      "grad_norm": 1.328125,
      "learning_rate": 8.981384345004732e-06,
      "loss": 0.9189,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 56
    },
    {
      "epoch": 0.14357682619647355,
      "grad_norm": 1.5078125,
      "learning_rate": 8.980265295073092e-06,
      "loss": 0.8623,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 57
    },
    {
      "epoch": 0.14609571788413098,
      "grad_norm": 1.203125,
      "learning_rate": 8.979143515381489e-06,
      "loss": 0.8184,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 58
    },
    {
      "epoch": 0.1486146095717884,
      "grad_norm": 1.453125,
      "learning_rate": 8.978018995929444e-06,
      "loss": 0.8418,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 59
    },
    {
      "epoch": 0.15113350125944586,
      "grad_norm": 1.1953125,
      "learning_rate": 8.976891726667572e-06,
      "loss": 0.9297,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 60
    },
    {
      "epoch": 0.15365239294710328,
      "grad_norm": 1.1953125,
      "learning_rate": 8.97576169749728e-06,
      "loss": 0.8945,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 61
    },
    {
      "epoch": 0.1561712846347607,
      "grad_norm": 1.578125,
      "learning_rate": 8.974628898270462e-06,
      "loss": 0.9258,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 62
    },
    {
      "epoch": 0.15869017632241814,
      "grad_norm": 1.234375,
      "learning_rate": 8.9734933187892e-06,
      "loss": 0.8389,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 63
    },
    {
      "epoch": 0.16120906801007556,
      "grad_norm": 1.1640625,
      "learning_rate": 8.97235494880546e-06,
      "loss": 0.8242,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 64
    },
    {
      "epoch": 0.163727959697733,
      "grad_norm": 2.734375,
      "learning_rate": 8.971213778020776e-06,
      "loss": 0.876,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 65
    },
    {
      "epoch": 0.16624685138539042,
      "grad_norm": 1.453125,
      "learning_rate": 8.970069796085946e-06,
      "loss": 0.9004,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 66
    },
    {
      "epoch": 0.16876574307304787,
      "grad_norm": 1.7421875,
      "learning_rate": 8.968922992600714e-06,
      "loss": 0.8809,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 67
    },
    {
      "epoch": 0.1712846347607053,
      "grad_norm": 1.390625,
      "learning_rate": 8.96777335711346e-06,
      "loss": 0.8721,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 68
    },
    {
      "epoch": 0.17380352644836272,
      "grad_norm": 1.5078125,
      "learning_rate": 8.966620879120879e-06,
      "loss": 0.8545,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 69
    },
    {
      "epoch": 0.17632241813602015,
      "grad_norm": 1.5859375,
      "learning_rate": 8.965465548067666e-06,
      "loss": 0.9199,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 70
    },
    {
      "epoch": 0.17884130982367757,
      "grad_norm": 1.234375,
      "learning_rate": 8.964307353346186e-06,
      "loss": 0.8467,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 71
    },
    {
      "epoch": 0.181360201511335,
      "grad_norm": 1.234375,
      "learning_rate": 8.963146284296154e-06,
      "loss": 0.9258,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 72
    },
    {
      "epoch": 0.18387909319899245,
      "grad_norm": 1.3125,
      "learning_rate": 8.961982330204308e-06,
      "loss": 0.8691,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 73
    },
    {
      "epoch": 0.18639798488664988,
      "grad_norm": 1.3046875,
      "learning_rate": 8.960815480304078e-06,
      "loss": 0.7891,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 74
    },
    {
      "epoch": 0.1889168765743073,
      "grad_norm": 1.3828125,
      "learning_rate": 8.959645723775257e-06,
      "loss": 0.8916,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 75
    },
    {
      "epoch": 0.19143576826196473,
      "grad_norm": 1.3515625,
      "learning_rate": 8.958473049743662e-06,
      "loss": 0.9062,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 76
    },
    {
      "epoch": 0.19395465994962216,
      "grad_norm": 1.21875,
      "learning_rate": 8.9572974472808e-06,
      "loss": 0.8076,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 77
    },
    {
      "epoch": 0.1964735516372796,
      "grad_norm": 1.2421875,
      "learning_rate": 8.956118905403529e-06,
      "loss": 0.874,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 78
    },
    {
      "epoch": 0.19899244332493704,
      "grad_norm": 1.3828125,
      "learning_rate": 8.954937413073714e-06,
      "loss": 0.9102,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 79
    },
    {
      "epoch": 0.20151133501259447,
      "grad_norm": 1.25,
      "learning_rate": 8.953752959197885e-06,
      "loss": 0.8408,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 80
    },
    {
      "epoch": 0.2040302267002519,
      "grad_norm": 1.3828125,
      "learning_rate": 8.952565532626881e-06,
      "loss": 0.958,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 81
    },
    {
      "epoch": 0.20654911838790932,
      "grad_norm": 1.421875,
      "learning_rate": 8.951375122155523e-06,
      "loss": 0.9668,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 82
    },
    {
      "epoch": 0.20906801007556675,
      "grad_norm": 1.2890625,
      "learning_rate": 8.950181716522227e-06,
      "loss": 0.8779,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 83
    },
    {
      "epoch": 0.21158690176322417,
      "grad_norm": 1.1796875,
      "learning_rate": 8.948985304408678e-06,
      "loss": 0.8057,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 84
    },
    {
      "epoch": 0.2141057934508816,
      "grad_norm": 1.203125,
      "learning_rate": 8.947785874439462e-06,
      "loss": 0.8672,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 85
    },
    {
      "epoch": 0.21662468513853905,
      "grad_norm": 1.5078125,
      "learning_rate": 8.946583415181705e-06,
      "loss": 0.8369,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 86
    },
    {
      "epoch": 0.21914357682619648,
      "grad_norm": 1.2265625,
      "learning_rate": 8.945377915144704e-06,
      "loss": 0.8311,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 87
    },
    {
      "epoch": 0.2216624685138539,
      "grad_norm": 1.1484375,
      "learning_rate": 8.944169362779576e-06,
      "loss": 0.8066,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 88
    },
    {
      "epoch": 0.22418136020151133,
      "grad_norm": 1.2109375,
      "learning_rate": 8.942957746478874e-06,
      "loss": 0.8584,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 89
    },
    {
      "epoch": 0.22670025188916876,
      "grad_norm": 1.203125,
      "learning_rate": 8.941743054576224e-06,
      "loss": 0.8223,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 90
    },
    {
      "epoch": 0.22921914357682618,
      "grad_norm": 1.1015625,
      "learning_rate": 8.940525275345949e-06,
      "loss": 0.8828,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 91
    },
    {
      "epoch": 0.23173803526448364,
      "grad_norm": 2.3125,
      "learning_rate": 8.939304397002686e-06,
      "loss": 0.9102,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 92
    },
    {
      "epoch": 0.23425692695214106,
      "grad_norm": 1.453125,
      "learning_rate": 8.938080407701019e-06,
      "loss": 0.8379,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 93
    },
    {
      "epoch": 0.2367758186397985,
      "grad_norm": 1.2421875,
      "learning_rate": 8.936853295535081e-06,
      "loss": 0.7842,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 94
    },
    {
      "epoch": 0.23929471032745592,
      "grad_norm": 1.2578125,
      "learning_rate": 8.935623048538179e-06,
      "loss": 0.7998,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 95
    },
    {
      "epoch": 0.24181360201511334,
      "grad_norm": 1.1796875,
      "learning_rate": 8.934389654682394e-06,
      "loss": 0.8838,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 96
    },
    {
      "epoch": 0.24433249370277077,
      "grad_norm": 1.1796875,
      "learning_rate": 8.933153101878202e-06,
      "loss": 0.8379,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 97
    },
    {
      "epoch": 0.24685138539042822,
      "grad_norm": 1.3828125,
      "learning_rate": 8.93191337797407e-06,
      "loss": 0.8896,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 98
    },
    {
      "epoch": 0.24937027707808565,
      "grad_norm": 1.1640625,
      "learning_rate": 8.930670470756064e-06,
      "loss": 0.8779,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 99
    },
    {
      "epoch": 0.2518891687657431,
      "grad_norm": 2.21875,
      "learning_rate": 8.929424367947436e-06,
      "loss": 0.8799,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 100
    },
    {
      "epoch": 0.25440806045340053,
      "grad_norm": 1.3671875,
      "learning_rate": 8.928175057208238e-06,
      "loss": 0.7812,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 101
    },
    {
      "epoch": 0.25692695214105793,
      "grad_norm": 1.5546875,
      "learning_rate": 8.926922526134899e-06,
      "loss": 0.7754,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 102
    },
    {
      "epoch": 0.2594458438287154,
      "grad_norm": 1.3359375,
      "learning_rate": 8.925666762259823e-06,
      "loss": 0.832,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 103
    },
    {
      "epoch": 0.2619647355163728,
      "grad_norm": 1.359375,
      "learning_rate": 8.924407753050969e-06,
      "loss": 0.8779,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 104
    },
    {
      "epoch": 0.26448362720403024,
      "grad_norm": 1.515625,
      "learning_rate": 8.923145485911444e-06,
      "loss": 0.7803,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 105
    },
    {
      "epoch": 0.26700251889168763,
      "grad_norm": 1.328125,
      "learning_rate": 8.92187994817907e-06,
      "loss": 0.9385,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 106
    },
    {
      "epoch": 0.2695214105793451,
      "grad_norm": 1.2421875,
      "learning_rate": 8.920611127125973e-06,
      "loss": 0.9365,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 107
    },
    {
      "epoch": 0.27204030226700254,
      "grad_norm": 1.234375,
      "learning_rate": 8.919339009958147e-06,
      "loss": 0.8379,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 108
    },
    {
      "epoch": 0.27455919395465994,
      "grad_norm": 1.1796875,
      "learning_rate": 8.918063583815029e-06,
      "loss": 0.8555,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 109
    },
    {
      "epoch": 0.2770780856423174,
      "grad_norm": 1.34375,
      "learning_rate": 8.916784835769064e-06,
      "loss": 0.9082,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 110
    },
    {
      "epoch": 0.2795969773299748,
      "grad_norm": 1.3046875,
      "learning_rate": 8.915502752825269e-06,
      "loss": 0.8721,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 111
    },
    {
      "epoch": 0.28211586901763225,
      "grad_norm": 1.1953125,
      "learning_rate": 8.914217321920789e-06,
      "loss": 0.8408,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 112
    },
    {
      "epoch": 0.28463476070528965,
      "grad_norm": 1.2421875,
      "learning_rate": 8.912928529924463e-06,
      "loss": 0.8203,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 113
    },
    {
      "epoch": 0.2871536523929471,
      "grad_norm": 1.734375,
      "learning_rate": 8.911636363636363e-06,
      "loss": 0.8457,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 114
    },
    {
      "epoch": 0.28967254408060455,
      "grad_norm": 1.3359375,
      "learning_rate": 8.910340809787358e-06,
      "loss": 0.8301,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 115
    },
    {
      "epoch": 0.29219143576826195,
      "grad_norm": 1.3203125,
      "learning_rate": 8.909041855038646e-06,
      "loss": 0.8018,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 116
    },
    {
      "epoch": 0.2947103274559194,
      "grad_norm": 1.2265625,
      "learning_rate": 8.907739485981309e-06,
      "loss": 0.7891,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 117
    },
    {
      "epoch": 0.2972292191435768,
      "grad_norm": 1.1015625,
      "learning_rate": 8.906433689135838e-06,
      "loss": 0.8418,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 118
    },
    {
      "epoch": 0.29974811083123426,
      "grad_norm": 1.453125,
      "learning_rate": 8.905124450951684e-06,
      "loss": 0.8604,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 119
    },
    {
      "epoch": 0.3022670025188917,
      "grad_norm": 2.21875,
      "learning_rate": 8.903811757806774e-06,
      "loss": 0.8984,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 120
    },
    {
      "epoch": 0.3047858942065491,
      "grad_norm": 1.5234375,
      "learning_rate": 8.902495596007047e-06,
      "loss": 0.9189,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 121
    },
    {
      "epoch": 0.30730478589420657,
      "grad_norm": 1.3515625,
      "learning_rate": 8.901175951785976e-06,
      "loss": 0.8018,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 122
    },
    {
      "epoch": 0.30982367758186397,
      "grad_norm": 1.265625,
      "learning_rate": 8.899852811304093e-06,
      "loss": 0.8066,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 123
    },
    {
      "epoch": 0.3123425692695214,
      "grad_norm": 1.171875,
      "learning_rate": 8.89852616064849e-06,
      "loss": 0.8506,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 124
    },
    {
      "epoch": 0.3148614609571788,
      "grad_norm": 1.0625,
      "learning_rate": 8.89719598583235e-06,
      "loss": 0.7969,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 125
    },
    {
      "epoch": 0.31738035264483627,
      "grad_norm": 1.421875,
      "learning_rate": 8.895862272794443e-06,
      "loss": 0.792,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 126
    },
    {
      "epoch": 0.3198992443324937,
      "grad_norm": 1.1171875,
      "learning_rate": 8.894525007398639e-06,
      "loss": 0.8096,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 127
    },
    {
      "epoch": 0.3224181360201511,
      "grad_norm": 1.2265625,
      "learning_rate": 8.893184175433398e-06,
      "loss": 0.7803,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 128
    },
    {
      "epoch": 0.3249370277078086,
      "grad_norm": 1.4453125,
      "learning_rate": 8.891839762611276e-06,
      "loss": 0.8662,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 129
    },
    {
      "epoch": 0.327455919395466,
      "grad_norm": 1.46875,
      "learning_rate": 8.890491754568414e-06,
      "loss": 0.7393,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 130
    },
    {
      "epoch": 0.32997481108312343,
      "grad_norm": 1.1875,
      "learning_rate": 8.889140136864028e-06,
      "loss": 0.7891,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 131
    },
    {
      "epoch": 0.33249370277078083,
      "grad_norm": 1.15625,
      "learning_rate": 8.88778489497989e-06,
      "loss": 0.9453,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 132
    },
    {
      "epoch": 0.3350125944584383,
      "grad_norm": 1.21875,
      "learning_rate": 8.88642601431981e-06,
      "loss": 0.7617,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 133
    },
    {
      "epoch": 0.33753148614609574,
      "grad_norm": 1.15625,
      "learning_rate": 8.88506348020911e-06,
      "loss": 0.8564,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 134
    },
    {
      "epoch": 0.34005037783375314,
      "grad_norm": 1.296875,
      "learning_rate": 8.883697277894106e-06,
      "loss": 0.8047,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 135
    },
    {
      "epoch": 0.3425692695214106,
      "grad_norm": 1.3046875,
      "learning_rate": 8.882327392541561e-06,
      "loss": 0.8721,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 136
    },
    {
      "epoch": 0.345088161209068,
      "grad_norm": 1.3046875,
      "learning_rate": 8.880953809238152e-06,
      "loss": 0.8652,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 137
    },
    {
      "epoch": 0.34760705289672544,
      "grad_norm": 1.3515625,
      "learning_rate": 8.879576512989938e-06,
      "loss": 0.8438,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 138
    },
    {
      "epoch": 0.3501259445843829,
      "grad_norm": 1.1015625,
      "learning_rate": 8.878195488721804e-06,
      "loss": 0.8008,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 139
    },
    {
      "epoch": 0.3526448362720403,
      "grad_norm": 1.53125,
      "learning_rate": 8.876810721276917e-06,
      "loss": 0.9053,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 140
    },
    {
      "epoch": 0.35516372795969775,
      "grad_norm": 1.3359375,
      "learning_rate": 8.875422195416164e-06,
      "loss": 0.8672,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 141
    },
    {
      "epoch": 0.35768261964735515,
      "grad_norm": 1.2265625,
      "learning_rate": 8.874029895817606e-06,
      "loss": 0.8008,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 142
    },
    {
      "epoch": 0.3602015113350126,
      "grad_norm": 1.1953125,
      "learning_rate": 8.8726338070759e-06,
      "loss": 0.7822,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 143
    },
    {
      "epoch": 0.36272040302267,
      "grad_norm": 1.140625,
      "learning_rate": 8.871233913701741e-06,
      "loss": 0.7842,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 144
    },
    {
      "epoch": 0.36523929471032746,
      "grad_norm": 1.1953125,
      "learning_rate": 8.869830200121285e-06,
      "loss": 0.7891,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 145
    },
    {
      "epoch": 0.3677581863979849,
      "grad_norm": 1.125,
      "learning_rate": 8.868422650675573e-06,
      "loss": 0.8408,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 146
    },
    {
      "epoch": 0.3702770780856423,
      "grad_norm": 1.2578125,
      "learning_rate": 8.867011249619946e-06,
      "loss": 0.8447,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 147
    },
    {
      "epoch": 0.37279596977329976,
      "grad_norm": 1.2421875,
      "learning_rate": 8.865595981123458e-06,
      "loss": 0.8613,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 148
    },
    {
      "epoch": 0.37531486146095716,
      "grad_norm": 1.4375,
      "learning_rate": 8.864176829268293e-06,
      "loss": 0.79,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 149
    },
    {
      "epoch": 0.3778337531486146,
      "grad_norm": 1.2578125,
      "learning_rate": 8.862753778049153e-06,
      "loss": 0.7529,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 150
    },
    {
      "epoch": 0.380352644836272,
      "grad_norm": 1.328125,
      "learning_rate": 8.86132681137267e-06,
      "loss": 0.8779,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 151
    },
    {
      "epoch": 0.38287153652392947,
      "grad_norm": 1.2890625,
      "learning_rate": 8.859895913056788e-06,
      "loss": 0.8652,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 152
    },
    {
      "epoch": 0.3853904282115869,
      "grad_norm": 1.265625,
      "learning_rate": 8.858461066830166e-06,
      "loss": 0.7529,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 153
    },
    {
      "epoch": 0.3879093198992443,
      "grad_norm": 1.359375,
      "learning_rate": 8.857022256331542e-06,
      "loss": 0.8027,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 154
    },
    {
      "epoch": 0.3904282115869018,
      "grad_norm": 1.40625,
      "learning_rate": 8.85557946510913e-06,
      "loss": 0.833,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 155
    },
    {
      "epoch": 0.3929471032745592,
      "grad_norm": 1.1328125,
      "learning_rate": 8.854132676619979e-06,
      "loss": 0.8008,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 156
    },
    {
      "epoch": 0.3954659949622166,
      "grad_norm": 1.1796875,
      "learning_rate": 8.852681874229348e-06,
      "loss": 0.8594,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 157
    },
    {
      "epoch": 0.3979848866498741,
      "grad_norm": 1.140625,
      "learning_rate": 8.851227041210064e-06,
      "loss": 0.7725,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 158
    },
    {
      "epoch": 0.4005037783375315,
      "grad_norm": 1.3828125,
      "learning_rate": 8.849768160741887e-06,
      "loss": 0.9307,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 159
    },
    {
      "epoch": 0.40302267002518893,
      "grad_norm": 1.296875,
      "learning_rate": 8.84830521591085e-06,
      "loss": 0.8184,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 160
    },
    {
      "epoch": 0.40554156171284633,
      "grad_norm": 1.0625,
      "learning_rate": 8.846838189708618e-06,
      "loss": 0.8447,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 161
    },
    {
      "epoch": 0.4080604534005038,
      "grad_norm": 1.1484375,
      "learning_rate": 8.845367065031817e-06,
      "loss": 0.8145,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 162
    },
    {
      "epoch": 0.4105793450881612,
      "grad_norm": 1.3671875,
      "learning_rate": 8.843891824681381e-06,
      "loss": 0.8037,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 163
    },
    {
      "epoch": 0.41309823677581864,
      "grad_norm": 1.125,
      "learning_rate": 8.842412451361868e-06,
      "loss": 0.832,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 164
    },
    {
      "epoch": 0.4156171284634761,
      "grad_norm": 1.1484375,
      "learning_rate": 8.840928927680798e-06,
      "loss": 0.8184,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 165
    },
    {
      "epoch": 0.4181360201511335,
      "grad_norm": 1.171875,
      "learning_rate": 8.839441236147964e-06,
      "loss": 0.8506,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 166
    },
    {
      "epoch": 0.42065491183879095,
      "grad_norm": 1.140625,
      "learning_rate": 8.837949359174743e-06,
      "loss": 0.8154,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 167
    },
    {
      "epoch": 0.42317380352644834,
      "grad_norm": 1.1015625,
      "learning_rate": 8.836453279073408e-06,
      "loss": 0.7705,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 168
    },
    {
      "epoch": 0.4256926952141058,
      "grad_norm": 1.34375,
      "learning_rate": 8.834952978056426e-06,
      "loss": 0.8115,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 169
    },
    {
      "epoch": 0.4282115869017632,
      "grad_norm": 1.421875,
      "learning_rate": 8.833448438235755e-06,
      "loss": 0.8203,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 170
    },
    {
      "epoch": 0.43073047858942065,
      "grad_norm": 1.25,
      "learning_rate": 8.831939641622132e-06,
      "loss": 0.7988,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 171
    },
    {
      "epoch": 0.4332493702770781,
      "grad_norm": 1.203125,
      "learning_rate": 8.83042657012435e-06,
      "loss": 0.7988,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 172
    },
    {
      "epoch": 0.4357682619647355,
      "grad_norm": 1.15625,
      "learning_rate": 8.82890920554855e-06,
      "loss": 0.8359,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 173
    },
    {
      "epoch": 0.43828715365239296,
      "grad_norm": 1.3515625,
      "learning_rate": 8.827387529597475e-06,
      "loss": 0.8623,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 174
    },
    {
      "epoch": 0.44080604534005036,
      "grad_norm": 1.4453125,
      "learning_rate": 8.825861523869745e-06,
      "loss": 0.8232,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 175
    },
    {
      "epoch": 0.4433249370277078,
      "grad_norm": 1.2734375,
      "learning_rate": 8.824331169859111e-06,
      "loss": 0.7861,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 176
    },
    {
      "epoch": 0.44584382871536526,
      "grad_norm": 1.21875,
      "learning_rate": 8.82279644895371e-06,
      "loss": 0.7861,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 177
    },
    {
      "epoch": 0.44836272040302266,
      "grad_norm": 1.3671875,
      "learning_rate": 8.821257342435307e-06,
      "loss": 0.7646,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 178
    },
    {
      "epoch": 0.4508816120906801,
      "grad_norm": 1.2734375,
      "learning_rate": 8.819713831478538e-06,
      "loss": 0.7959,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 179
    },
    {
      "epoch": 0.4534005037783375,
      "grad_norm": 1.4140625,
      "learning_rate": 8.818165897150136e-06,
      "loss": 0.8311,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 180
    },
    {
      "epoch": 0.45591939546599497,
      "grad_norm": 1.0859375,
      "learning_rate": 8.816613520408164e-06,
      "loss": 0.748,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 181
    },
    {
      "epoch": 0.45843828715365237,
      "grad_norm": 1.5625,
      "learning_rate": 8.815056682101229e-06,
      "loss": 0.8076,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 182
    },
    {
      "epoch": 0.4609571788413098,
      "grad_norm": 1.28125,
      "learning_rate": 8.8134953629677e-06,
      "loss": 0.7988,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 183
    },
    {
      "epoch": 0.4634760705289673,
      "grad_norm": 1.3671875,
      "learning_rate": 8.81192954363491e-06,
      "loss": 0.8828,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 184
    },
    {
      "epoch": 0.4659949622166247,
      "grad_norm": 1.234375,
      "learning_rate": 8.810359204618346e-06,
      "loss": 0.8242,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 185
    },
    {
      "epoch": 0.46851385390428213,
      "grad_norm": 1.640625,
      "learning_rate": 8.808784326320862e-06,
      "loss": 0.7812,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 186
    },
    {
      "epoch": 0.47103274559193953,
      "grad_norm": 1.0859375,
      "learning_rate": 8.807204889031843e-06,
      "loss": 0.7666,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 187
    },
    {
      "epoch": 0.473551637279597,
      "grad_norm": 1.296875,
      "learning_rate": 8.805620872926398e-06,
      "loss": 0.7598,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 188
    },
    {
      "epoch": 0.4760705289672544,
      "grad_norm": 1.046875,
      "learning_rate": 8.804032258064517e-06,
      "loss": 0.7295,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 189
    },
    {
      "epoch": 0.47858942065491183,
      "grad_norm": 1.3125,
      "learning_rate": 8.802439024390243e-06,
      "loss": 0.8799,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 190
    },
    {
      "epoch": 0.4811083123425693,
      "grad_norm": 1.140625,
      "learning_rate": 8.80084115173083e-06,
      "loss": 0.7549,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 191
    },
    {
      "epoch": 0.4836272040302267,
      "grad_norm": 1.265625,
      "learning_rate": 8.799238619795886e-06,
      "loss": 0.8633,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 192
    },
    {
      "epoch": 0.48614609571788414,
      "grad_norm": 1.109375,
      "learning_rate": 8.79763140817651e-06,
      "loss": 0.8408,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 193
    },
    {
      "epoch": 0.48866498740554154,
      "grad_norm": 1.2109375,
      "learning_rate": 8.796019496344436e-06,
      "loss": 0.8496,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 194
    },
    {
      "epoch": 0.491183879093199,
      "grad_norm": 1.671875,
      "learning_rate": 8.794402863651155e-06,
      "loss": 0.833,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 195
    },
    {
      "epoch": 0.49370277078085645,
      "grad_norm": 1.421875,
      "learning_rate": 8.792781489327033e-06,
      "loss": 0.8457,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 196
    },
    {
      "epoch": 0.49622166246851385,
      "grad_norm": 1.109375,
      "learning_rate": 8.791155352480418e-06,
      "loss": 0.7988,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 197
    },
    {
      "epoch": 0.4987405541561713,
      "grad_norm": 1.2265625,
      "learning_rate": 8.789524432096748e-06,
      "loss": 0.8125,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 198
    },
    {
      "epoch": 0.5012594458438288,
      "grad_norm": 1.1953125,
      "learning_rate": 8.787888707037644e-06,
      "loss": 0.7598,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 199
    },
    {
      "epoch": 0.5037783375314862,
      "grad_norm": 1.2109375,
      "learning_rate": 8.786248156039995e-06,
      "loss": 0.8047,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 200
    },
    {
      "epoch": 0.5062972292191436,
      "grad_norm": 1.140625,
      "learning_rate": 8.784602757715037e-06,
      "loss": 0.8418,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 201
    },
    {
      "epoch": 0.5088161209068011,
      "grad_norm": 1.3125,
      "learning_rate": 8.782952490547428e-06,
      "loss": 0.8369,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 202
    },
    {
      "epoch": 0.5113350125944585,
      "grad_norm": 2.515625,
      "learning_rate": 8.781297332894304e-06,
      "loss": 0.79,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 203
    },
    {
      "epoch": 0.5138539042821159,
      "grad_norm": 1.4453125,
      "learning_rate": 8.779637262984337e-06,
      "loss": 0.8691,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 204
    },
    {
      "epoch": 0.5163727959697733,
      "grad_norm": 1.4765625,
      "learning_rate": 8.777972258916777e-06,
      "loss": 0.8809,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 205
    },
    {
      "epoch": 0.5188916876574308,
      "grad_norm": 1.2421875,
      "learning_rate": 8.776302298660493e-06,
      "loss": 0.8086,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 206
    },
    {
      "epoch": 0.5214105793450882,
      "grad_norm": 1.1484375,
      "learning_rate": 8.774627360052997e-06,
      "loss": 0.7715,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 207
    },
    {
      "epoch": 0.5239294710327456,
      "grad_norm": 1.2734375,
      "learning_rate": 8.77294742079947e-06,
      "loss": 0.8145,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 208
    },
    {
      "epoch": 0.5264483627204031,
      "grad_norm": 1.34375,
      "learning_rate": 8.771262458471761e-06,
      "loss": 0.7949,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 209
    },
    {
      "epoch": 0.5289672544080605,
      "grad_norm": 1.3671875,
      "learning_rate": 8.769572450507402e-06,
      "loss": 0.75,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 210
    },
    {
      "epoch": 0.5314861460957179,
      "grad_norm": 1.203125,
      "learning_rate": 8.767877374208597e-06,
      "loss": 0.8477,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 211
    },
    {
      "epoch": 0.5340050377833753,
      "grad_norm": 1.171875,
      "learning_rate": 8.7661772067412e-06,
      "loss": 0.8154,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 212
    },
    {
      "epoch": 0.5365239294710328,
      "grad_norm": 1.125,
      "learning_rate": 8.76447192513369e-06,
      "loss": 0.7559,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 213
    },
    {
      "epoch": 0.5390428211586902,
      "grad_norm": 1.2265625,
      "learning_rate": 8.76276150627615e-06,
      "loss": 0.8047,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 214
    },
    {
      "epoch": 0.5415617128463476,
      "grad_norm": 1.1015625,
      "learning_rate": 8.761045926919207e-06,
      "loss": 0.792,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 215
    },
    {
      "epoch": 0.5440806045340051,
      "grad_norm": 1.1640625,
      "learning_rate": 8.75932516367299e-06,
      "loss": 0.7939,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 216
    },
    {
      "epoch": 0.5465994962216625,
      "grad_norm": 1.21875,
      "learning_rate": 8.757599193006053e-06,
      "loss": 0.7637,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 217
    },
    {
      "epoch": 0.5491183879093199,
      "grad_norm": 1.2421875,
      "learning_rate": 8.755867991244318e-06,
      "loss": 0.7939,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 218
    },
    {
      "epoch": 0.5516372795969773,
      "grad_norm": 1.0546875,
      "learning_rate": 8.754131534569984e-06,
      "loss": 0.7812,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 219
    },
    {
      "epoch": 0.5541561712846348,
      "grad_norm": 1.203125,
      "learning_rate": 8.752389799020436e-06,
      "loss": 0.7695,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 220
    },
    {
      "epoch": 0.5566750629722922,
      "grad_norm": 1.2265625,
      "learning_rate": 8.750642760487145e-06,
      "loss": 0.8857,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 221
    },
    {
      "epoch": 0.5591939546599496,
      "grad_norm": 1.1953125,
      "learning_rate": 8.748890394714553e-06,
      "loss": 0.8574,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 222
    },
    {
      "epoch": 0.5617128463476071,
      "grad_norm": 1.125,
      "learning_rate": 8.747132677298948e-06,
      "loss": 0.8193,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 223
    },
    {
      "epoch": 0.5642317380352645,
      "grad_norm": 1.0234375,
      "learning_rate": 8.74536958368734e-06,
      "loss": 0.7441,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 224
    },
    {
      "epoch": 0.5667506297229219,
      "grad_norm": 1.203125,
      "learning_rate": 8.74360108917631e-06,
      "loss": 0.792,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 225
    },
    {
      "epoch": 0.5692695214105793,
      "grad_norm": 1.28125,
      "learning_rate": 8.741827168910857e-06,
      "loss": 0.8105,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 226
    },
    {
      "epoch": 0.5717884130982368,
      "grad_norm": 1.0625,
      "learning_rate": 8.740047797883237e-06,
      "loss": 0.8115,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 227
    },
    {
      "epoch": 0.5743073047858942,
      "grad_norm": 1.140625,
      "learning_rate": 8.738262950931784e-06,
      "loss": 0.7998,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 228
    },
    {
      "epoch": 0.5768261964735516,
      "grad_norm": 1.234375,
      "learning_rate": 8.736472602739726e-06,
      "loss": 0.7686,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 229
    },
    {
      "epoch": 0.5793450881612091,
      "grad_norm": 1.203125,
      "learning_rate": 8.73467672783399e-06,
      "loss": 0.7959,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 230
    },
    {
      "epoch": 0.5818639798488665,
      "grad_norm": 1.1796875,
      "learning_rate": 8.73287530058399e-06,
      "loss": 0.6982,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 231
    },
    {
      "epoch": 0.5843828715365239,
      "grad_norm": 1.21875,
      "learning_rate": 8.731068295200414e-06,
      "loss": 0.7695,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 232
    },
    {
      "epoch": 0.5869017632241813,
      "grad_norm": 1.1953125,
      "learning_rate": 8.729255685733977e-06,
      "loss": 0.7637,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 233
    },
    {
      "epoch": 0.5894206549118388,
      "grad_norm": 1.15625,
      "learning_rate": 8.727437446074203e-06,
      "loss": 0.7646,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 234
    },
    {
      "epoch": 0.5919395465994962,
      "grad_norm": 1.125,
      "learning_rate": 8.72561354994815e-06,
      "loss": 0.7607,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 235
    },
    {
      "epoch": 0.5944584382871536,
      "grad_norm": 10.6875,
      "learning_rate": 8.723783970919162e-06,
      "loss": 0.8115,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 236
    },
    {
      "epoch": 0.5969773299748111,
      "grad_norm": 1.171875,
      "learning_rate": 8.721948682385575e-06,
      "loss": 0.7295,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 237
    },
    {
      "epoch": 0.5994962216624685,
      "grad_norm": 1.1953125,
      "learning_rate": 8.720107657579442e-06,
      "loss": 0.7764,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 238
    },
    {
      "epoch": 0.6020151133501259,
      "grad_norm": 1.09375,
      "learning_rate": 8.718260869565218e-06,
      "loss": 0.8428,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 239
    },
    {
      "epoch": 0.6045340050377834,
      "grad_norm": 1.1796875,
      "learning_rate": 8.716408291238462e-06,
      "loss": 0.7832,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 240
    },
    {
      "epoch": 0.6070528967254408,
      "grad_norm": 1.171875,
      "learning_rate": 8.714549895324495e-06,
      "loss": 0.7861,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 241
    },
    {
      "epoch": 0.6095717884130982,
      "grad_norm": 1.2109375,
      "learning_rate": 8.712685654377075e-06,
      "loss": 0.7676,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 242
    },
    {
      "epoch": 0.6120906801007556,
      "grad_norm": 1.15625,
      "learning_rate": 8.710815540777039e-06,
      "loss": 0.7598,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 243
    },
    {
      "epoch": 0.6146095717884131,
      "grad_norm": 1.140625,
      "learning_rate": 8.708939526730937e-06,
      "loss": 0.7979,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 244
    },
    {
      "epoch": 0.6171284634760705,
      "grad_norm": 1.2421875,
      "learning_rate": 8.707057584269663e-06,
      "loss": 0.7959,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 245
    },
    {
      "epoch": 0.6196473551637279,
      "grad_norm": 1.125,
      "learning_rate": 8.705169685247055e-06,
      "loss": 0.8115,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 246
    },
    {
      "epoch": 0.6221662468513854,
      "grad_norm": 1.0546875,
      "learning_rate": 8.7032758013385e-06,
      "loss": 0.7598,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 247
    },
    {
      "epoch": 0.6246851385390428,
      "grad_norm": 1.1953125,
      "learning_rate": 8.701375904039512e-06,
      "loss": 0.8301,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 248
    },
    {
      "epoch": 0.6272040302267002,
      "grad_norm": 1.609375,
      "learning_rate": 8.699469964664311e-06,
      "loss": 0.7529,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 249
    },
    {
      "epoch": 0.6297229219143576,
      "grad_norm": 1.203125,
      "learning_rate": 8.697557954344365e-06,
      "loss": 0.7852,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 250
    },
    {
      "epoch": 0.6322418136020151,
      "grad_norm": 1.1953125,
      "learning_rate": 8.695639844026941e-06,
      "loss": 0.8721,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 251
    },
    {
      "epoch": 0.6347607052896725,
      "grad_norm": 1.2421875,
      "learning_rate": 8.693715604473637e-06,
      "loss": 0.915,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 252
    },
    {
      "epoch": 0.6372795969773299,
      "grad_norm": 1.1640625,
      "learning_rate": 8.691785206258892e-06,
      "loss": 0.752,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 253
    },
    {
      "epoch": 0.6397984886649875,
      "grad_norm": 1.546875,
      "learning_rate": 8.689848619768477e-06,
      "loss": 0.8145,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 254
    },
    {
      "epoch": 0.6423173803526449,
      "grad_norm": 1.296875,
      "learning_rate": 8.687905815198002e-06,
      "loss": 0.8467,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 255
    },
    {
      "epoch": 0.6448362720403022,
      "grad_norm": 1.328125,
      "learning_rate": 8.685956762551366e-06,
      "loss": 0.8008,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 256
    },
    {
      "epoch": 0.6473551637279596,
      "grad_norm": 1.140625,
      "learning_rate": 8.684001431639228e-06,
      "loss": 0.7744,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 257
    },
    {
      "epoch": 0.6498740554156172,
      "grad_norm": 1.03125,
      "learning_rate": 8.682039792077434e-06,
      "loss": 0.7354,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 258
    },
    {
      "epoch": 0.6523929471032746,
      "grad_norm": 1.1171875,
      "learning_rate": 8.680071813285458e-06,
      "loss": 0.833,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 259
    },
    {
      "epoch": 0.654911838790932,
      "grad_norm": 1.1640625,
      "learning_rate": 8.678097464484805e-06,
      "loss": 0.7764,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 260
    },
    {
      "epoch": 0.6574307304785895,
      "grad_norm": 1.46875,
      "learning_rate": 8.676116714697407e-06,
      "loss": 0.8633,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 261
    },
    {
      "epoch": 0.6599496221662469,
      "grad_norm": 1.09375,
      "learning_rate": 8.674129532744001e-06,
      "loss": 0.8398,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 262
    },
    {
      "epoch": 0.6624685138539043,
      "grad_norm": 1.3671875,
      "learning_rate": 8.672135887242501e-06,
      "loss": 0.8594,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 263
    },
    {
      "epoch": 0.6649874055415617,
      "grad_norm": 1.234375,
      "learning_rate": 8.670135746606335e-06,
      "loss": 0.7314,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 264
    },
    {
      "epoch": 0.6675062972292192,
      "grad_norm": 1.0546875,
      "learning_rate": 8.668129079042786e-06,
      "loss": 0.7266,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 265
    },
    {
      "epoch": 0.6700251889168766,
      "grad_norm": 1.2734375,
      "learning_rate": 8.666115852551298e-06,
      "loss": 0.9062,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 266
    },
    {
      "epoch": 0.672544080604534,
      "grad_norm": 1.1796875,
      "learning_rate": 8.66409603492179e-06,
      "loss": 0.707,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 267
    },
    {
      "epoch": 0.6750629722921915,
      "grad_norm": 1.046875,
      "learning_rate": 8.66206959373292e-06,
      "loss": 0.7881,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 268
    },
    {
      "epoch": 0.6775818639798489,
      "grad_norm": 1.1328125,
      "learning_rate": 8.660036496350365e-06,
      "loss": 0.7598,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 269
    },
    {
      "epoch": 0.6801007556675063,
      "grad_norm": 1.3125,
      "learning_rate": 8.657996709925059e-06,
      "loss": 0.8418,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 270
    },
    {
      "epoch": 0.6826196473551638,
      "grad_norm": 1.2421875,
      "learning_rate": 8.655950201391432e-06,
      "loss": 0.7588,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 271
    },
    {
      "epoch": 0.6851385390428212,
      "grad_norm": 1.0625,
      "learning_rate": 8.653896937465615e-06,
      "loss": 0.7383,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 272
    },
    {
      "epoch": 0.6876574307304786,
      "grad_norm": 1.28125,
      "learning_rate": 8.651836884643645e-06,
      "loss": 0.7686,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 273
    },
    {
      "epoch": 0.690176322418136,
      "grad_norm": 1.328125,
      "learning_rate": 8.649770009199633e-06,
      "loss": 0.8057,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 274
    },
    {
      "epoch": 0.6926952141057935,
      "grad_norm": 1.28125,
      "learning_rate": 8.647696277183928e-06,
      "loss": 0.79,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 275
    },
    {
      "epoch": 0.6952141057934509,
      "grad_norm": 1.1171875,
      "learning_rate": 8.645615654421265e-06,
      "loss": 0.8838,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 276
    },
    {
      "epoch": 0.6977329974811083,
      "grad_norm": 1.5703125,
      "learning_rate": 8.643528106508877e-06,
      "loss": 0.7314,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 277
    },
    {
      "epoch": 0.7002518891687658,
      "grad_norm": 1.3359375,
      "learning_rate": 8.641433598814596e-06,
      "loss": 0.7676,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 278
    },
    {
      "epoch": 0.7027707808564232,
      "grad_norm": 1.2890625,
      "learning_rate": 8.639332096474953e-06,
      "loss": 0.8281,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 279
    },
    {
      "epoch": 0.7052896725440806,
      "grad_norm": 1.15625,
      "learning_rate": 8.637223564393235e-06,
      "loss": 0.7646,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 280
    },
    {
      "epoch": 0.707808564231738,
      "grad_norm": 1.2265625,
      "learning_rate": 8.635107967237528e-06,
      "loss": 0.8105,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 281
    },
    {
      "epoch": 0.7103274559193955,
      "grad_norm": 1.2265625,
      "learning_rate": 8.632985269438747e-06,
      "loss": 0.8438,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 282
    },
    {
      "epoch": 0.7128463476070529,
      "grad_norm": 1.109375,
      "learning_rate": 8.630855435188644e-06,
      "loss": 0.7891,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 283
    },
    {
      "epoch": 0.7153652392947103,
      "grad_norm": 1.0859375,
      "learning_rate": 8.628718428437793e-06,
      "loss": 0.9287,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 284
    },
    {
      "epoch": 0.7178841309823678,
      "grad_norm": 1.125,
      "learning_rate": 8.626574212893554e-06,
      "loss": 0.7598,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 285
    },
    {
      "epoch": 0.7204030226700252,
      "grad_norm": 1.140625,
      "learning_rate": 8.624422752018023e-06,
      "loss": 0.8291,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 286
    },
    {
      "epoch": 0.7229219143576826,
      "grad_norm": 1.296875,
      "learning_rate": 8.62226400902595e-06,
      "loss": 0.7832,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 287
    },
    {
      "epoch": 0.72544080604534,
      "grad_norm": 1.6328125,
      "learning_rate": 8.62009794688265e-06,
      "loss": 0.8164,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 288
    },
    {
      "epoch": 0.7279596977329975,
      "grad_norm": 1.234375,
      "learning_rate": 8.617924528301888e-06,
      "loss": 0.7773,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 289
    },
    {
      "epoch": 0.7304785894206549,
      "grad_norm": 1.171875,
      "learning_rate": 8.615743715743717e-06,
      "loss": 0.8291,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 290
    },
    {
      "epoch": 0.7329974811083123,
      "grad_norm": 1.078125,
      "learning_rate": 8.613555471412343e-06,
      "loss": 0.7959,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 291
    },
    {
      "epoch": 0.7355163727959698,
      "grad_norm": 1.40625,
      "learning_rate": 8.611359757253934e-06,
      "loss": 0.7549,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 292
    },
    {
      "epoch": 0.7380352644836272,
      "grad_norm": 1.0703125,
      "learning_rate": 8.609156534954409e-06,
      "loss": 0.7715,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 293
    },
    {
      "epoch": 0.7405541561712846,
      "grad_norm": 1.1015625,
      "learning_rate": 8.606945765937203e-06,
      "loss": 0.7861,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 294
    },
    {
      "epoch": 0.743073047858942,
      "grad_norm": 1.3046875,
      "learning_rate": 8.604727411361037e-06,
      "loss": 0.8574,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 295
    },
    {
      "epoch": 0.7455919395465995,
      "grad_norm": 1.125,
      "learning_rate": 8.602501432117624e-06,
      "loss": 0.75,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 296
    },
    {
      "epoch": 0.7481108312342569,
      "grad_norm": 1.0546875,
      "learning_rate": 8.600267788829381e-06,
      "loss": 0.8682,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 297
    },
    {
      "epoch": 0.7506297229219143,
      "grad_norm": 1.078125,
      "learning_rate": 8.598026441847098e-06,
      "loss": 0.8125,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 298
    },
    {
      "epoch": 0.7531486146095718,
      "grad_norm": 1.09375,
      "learning_rate": 8.595777351247601e-06,
      "loss": 0.8555,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 299
    },
    {
      "epoch": 0.7556675062972292,
      "grad_norm": 1.15625,
      "learning_rate": 8.593520476831378e-06,
      "loss": 0.7256,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 300
    },
    {
      "epoch": 0.7581863979848866,
      "grad_norm": 1.0546875,
      "learning_rate": 8.591255778120185e-06,
      "loss": 0.8066,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 301
    },
    {
      "epoch": 0.760705289672544,
      "grad_norm": 1.1015625,
      "learning_rate": 8.58898321435462e-06,
      "loss": 0.8301,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 302
    },
    {
      "epoch": 0.7632241813602015,
      "grad_norm": 1.1171875,
      "learning_rate": 8.58670274449169e-06,
      "loss": 0.7686,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 303
    },
    {
      "epoch": 0.7657430730478589,
      "grad_norm": 1.234375,
      "learning_rate": 8.584414327202324e-06,
      "loss": 0.7969,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 304
    },
    {
      "epoch": 0.7682619647355163,
      "grad_norm": 1.171875,
      "learning_rate": 8.582117920868892e-06,
      "loss": 0.8574,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 305
    },
    {
      "epoch": 0.7707808564231738,
      "grad_norm": 1.0703125,
      "learning_rate": 8.57981348358267e-06,
      "loss": 0.7412,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 306
    },
    {
      "epoch": 0.7732997481108312,
      "grad_norm": 1.265625,
      "learning_rate": 8.5775009731413e-06,
      "loss": 0.9023,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 307
    },
    {
      "epoch": 0.7758186397984886,
      "grad_norm": 1.21875,
      "learning_rate": 8.575180347046208e-06,
      "loss": 0.8477,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 308
    },
    {
      "epoch": 0.7783375314861462,
      "grad_norm": 1.1015625,
      "learning_rate": 8.5728515625e-06,
      "loss": 0.8184,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 309
    },
    {
      "epoch": 0.7808564231738035,
      "grad_norm": 1.0078125,
      "learning_rate": 8.570514576403836e-06,
      "loss": 0.7373,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 310
    },
    {
      "epoch": 0.783375314861461,
      "grad_norm": 1.1875,
      "learning_rate": 8.568169345354762e-06,
      "loss": 0.7637,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 311
    },
    {
      "epoch": 0.7858942065491183,
      "grad_norm": 1.1328125,
      "learning_rate": 8.565815825643039e-06,
      "loss": 0.7979,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 312
    },
    {
      "epoch": 0.7884130982367759,
      "grad_norm": 1.3203125,
      "learning_rate": 8.56345397324941e-06,
      "loss": 0.8418,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 313
    },
    {
      "epoch": 0.7909319899244333,
      "grad_norm": 1.1796875,
      "learning_rate": 8.561083743842366e-06,
      "loss": 0.8135,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 314
    },
    {
      "epoch": 0.7934508816120907,
      "grad_norm": 1.1953125,
      "learning_rate": 8.558705092775365e-06,
      "loss": 0.8242,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 315
    },
    {
      "epoch": 0.7959697732997482,
      "grad_norm": 1.34375,
      "learning_rate": 8.556317975084042e-06,
      "loss": 0.6963,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 316
    },
    {
      "epoch": 0.7984886649874056,
      "grad_norm": 1.453125,
      "learning_rate": 8.55392234548336e-06,
      "loss": 0.8691,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 317
    },
    {
      "epoch": 0.801007556675063,
      "grad_norm": 1.078125,
      "learning_rate": 8.551518158364756e-06,
      "loss": 0.7109,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 318
    },
    {
      "epoch": 0.8035264483627204,
      "grad_norm": 1.15625,
      "learning_rate": 8.54910536779324e-06,
      "loss": 0.7812,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 319
    },
    {
      "epoch": 0.8060453400503779,
      "grad_norm": 1.1171875,
      "learning_rate": 8.546683927504481e-06,
      "loss": 0.7041,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 320
    },
    {
      "epoch": 0.8085642317380353,
      "grad_norm": 1.2109375,
      "learning_rate": 8.544253790901836e-06,
      "loss": 0.8018,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 321
    },
    {
      "epoch": 0.8110831234256927,
      "grad_norm": 1.0703125,
      "learning_rate": 8.54181491105337e-06,
      "loss": 0.7627,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 322
    },
    {
      "epoch": 0.8136020151133502,
      "grad_norm": 1.171875,
      "learning_rate": 8.539367240688826e-06,
      "loss": 0.75,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 323
    },
    {
      "epoch": 0.8161209068010076,
      "grad_norm": 1.15625,
      "learning_rate": 8.536910732196591e-06,
      "loss": 0.7305,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 324
    },
    {
      "epoch": 0.818639798488665,
      "grad_norm": 1.0703125,
      "learning_rate": 8.534445337620579e-06,
      "loss": 0.7949,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 325
    },
    {
      "epoch": 0.8211586901763224,
      "grad_norm": 1.1171875,
      "learning_rate": 8.531971008657139e-06,
      "loss": 0.79,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 326
    },
    {
      "epoch": 0.8236775818639799,
      "grad_norm": 1.0703125,
      "learning_rate": 8.529487696651876e-06,
      "loss": 0.7666,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 327
    },
    {
      "epoch": 0.8261964735516373,
      "grad_norm": 1.0078125,
      "learning_rate": 8.526995352596485e-06,
      "loss": 0.6963,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 328
    },
    {
      "epoch": 0.8287153652392947,
      "grad_norm": 1.0390625,
      "learning_rate": 8.524493927125505e-06,
      "loss": 0.7451,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 329
    },
    {
      "epoch": 0.8312342569269522,
      "grad_norm": 1.078125,
      "learning_rate": 8.521983370513081e-06,
      "loss": 0.7168,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 330
    },
    {
      "epoch": 0.8337531486146096,
      "grad_norm": 1.5390625,
      "learning_rate": 8.519463632669648e-06,
      "loss": 0.7754,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 331
    },
    {
      "epoch": 0.836272040302267,
      "grad_norm": 1.265625,
      "learning_rate": 8.51693466313861e-06,
      "loss": 0.8301,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 332
    },
    {
      "epoch": 0.8387909319899244,
      "grad_norm": 1.3046875,
      "learning_rate": 8.514396411092986e-06,
      "loss": 0.7988,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 333
    },
    {
      "epoch": 0.8413098236775819,
      "grad_norm": 3.15625,
      "learning_rate": 8.511848825331972e-06,
      "loss": 0.7119,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 334
    },
    {
      "epoch": 0.8438287153652393,
      "grad_norm": 1.1171875,
      "learning_rate": 8.509291854277527e-06,
      "loss": 0.7715,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 335
    },
    {
      "epoch": 0.8463476070528967,
      "grad_norm": 2.578125,
      "learning_rate": 8.506725445970883e-06,
      "loss": 0.7773,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 336
    },
    {
      "epoch": 0.8488664987405542,
      "grad_norm": 1.21875,
      "learning_rate": 8.504149548069022e-06,
      "loss": 0.7607,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 337
    },
    {
      "epoch": 0.8513853904282116,
      "grad_norm": 1.2109375,
      "learning_rate": 8.501564107841121e-06,
      "loss": 0.752,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 338
    },
    {
      "epoch": 0.853904282115869,
      "grad_norm": 1.1328125,
      "learning_rate": 8.498969072164949e-06,
      "loss": 0.7812,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 339
    },
    {
      "epoch": 0.8564231738035264,
      "grad_norm": 1.1171875,
      "learning_rate": 8.496364387523239e-06,
      "loss": 0.8232,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 340
    },
    {
      "epoch": 0.8589420654911839,
      "grad_norm": 1.234375,
      "learning_rate": 8.49375e-06,
      "loss": 0.8281,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 341
    },
    {
      "epoch": 0.8614609571788413,
      "grad_norm": 1.2734375,
      "learning_rate": 8.4911258552768e-06,
      "loss": 0.9336,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 342
    },
    {
      "epoch": 0.8639798488664987,
      "grad_norm": 1.4921875,
      "learning_rate": 8.488491898629e-06,
      "loss": 0.7705,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 343
    },
    {
      "epoch": 0.8664987405541562,
      "grad_norm": 1.359375,
      "learning_rate": 8.485848074921957e-06,
      "loss": 0.8467,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 344
    },
    {
      "epoch": 0.8690176322418136,
      "grad_norm": 1.1171875,
      "learning_rate": 8.483194328607173e-06,
      "loss": 0.7803,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 345
    },
    {
      "epoch": 0.871536523929471,
      "grad_norm": 1.1953125,
      "learning_rate": 8.480530603718405e-06,
      "loss": 0.792,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 346
    },
    {
      "epoch": 0.8740554156171285,
      "grad_norm": 1.1328125,
      "learning_rate": 8.477856843867728e-06,
      "loss": 0.832,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 347
    },
    {
      "epoch": 0.8765743073047859,
      "grad_norm": 1.0078125,
      "learning_rate": 8.47517299224156e-06,
      "loss": 0.7949,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 348
    },
    {
      "epoch": 0.8790931989924433,
      "grad_norm": 1.078125,
      "learning_rate": 8.472478991596638e-06,
      "loss": 0.7598,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 349
    },
    {
      "epoch": 0.8816120906801007,
      "grad_norm": 1.09375,
      "learning_rate": 8.469774784255946e-06,
      "loss": 0.7158,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 350
    },
    {
      "epoch": 0.8841309823677582,
      "grad_norm": 1.3125,
      "learning_rate": 8.467060312104597e-06,
      "loss": 0.7334,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 351
    },
    {
      "epoch": 0.8866498740554156,
      "grad_norm": 1.1484375,
      "learning_rate": 8.464335516585674e-06,
      "loss": 0.8545,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 352
    },
    {
      "epoch": 0.889168765743073,
      "grad_norm": 1.21875,
      "learning_rate": 8.46160033869602e-06,
      "loss": 0.748,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 353
    },
    {
      "epoch": 0.8916876574307305,
      "grad_norm": 1.265625,
      "learning_rate": 8.458854718981973e-06,
      "loss": 0.7686,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 354
    },
    {
      "epoch": 0.8942065491183879,
      "grad_norm": 1.25,
      "learning_rate": 8.456098597535062e-06,
      "loss": 0.8613,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 355
    },
    {
      "epoch": 0.8967254408060453,
      "grad_norm": 1.1484375,
      "learning_rate": 8.453331913987651e-06,
      "loss": 0.7715,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 356
    },
    {
      "epoch": 0.8992443324937027,
      "grad_norm": 1.1328125,
      "learning_rate": 8.450554607508531e-06,
      "loss": 0.7793,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 357
    },
    {
      "epoch": 0.9017632241813602,
      "grad_norm": 1.15625,
      "learning_rate": 8.447766616798463e-06,
      "loss": 0.8506,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 358
    },
    {
      "epoch": 0.9042821158690176,
      "grad_norm": 0.984375,
      "learning_rate": 8.444967880085653e-06,
      "loss": 0.7002,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 359
    },
    {
      "epoch": 0.906801007556675,
      "grad_norm": 1.0625,
      "learning_rate": 8.442158335121218e-06,
      "loss": 0.8105,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 360
    },
    {
      "epoch": 0.9093198992443325,
      "grad_norm": 1.1953125,
      "learning_rate": 8.439337919174549e-06,
      "loss": 0.751,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 361
    },
    {
      "epoch": 0.9118387909319899,
      "grad_norm": 1.171875,
      "learning_rate": 8.436506569028646e-06,
      "loss": 0.8604,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 362
    },
    {
      "epoch": 0.9143576826196473,
      "grad_norm": 1.3828125,
      "learning_rate": 8.4336642209754e-06,
      "loss": 0.8877,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 363
    },
    {
      "epoch": 0.9168765743073047,
      "grad_norm": 1.046875,
      "learning_rate": 8.430810810810811e-06,
      "loss": 0.79,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 364
    },
    {
      "epoch": 0.9193954659949622,
      "grad_norm": 1.0703125,
      "learning_rate": 8.427946273830157e-06,
      "loss": 0.7539,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 365
    },
    {
      "epoch": 0.9219143576826196,
      "grad_norm": 1.21875,
      "learning_rate": 8.425070544823096e-06,
      "loss": 0.7451,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 366
    },
    {
      "epoch": 0.924433249370277,
      "grad_norm": 1.0625,
      "learning_rate": 8.422183558068725e-06,
      "loss": 0.8232,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 367
    },
    {
      "epoch": 0.9269521410579346,
      "grad_norm": 1.234375,
      "learning_rate": 8.419285247330573e-06,
      "loss": 0.7607,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 368
    },
    {
      "epoch": 0.929471032745592,
      "grad_norm": 1.1484375,
      "learning_rate": 8.416375545851529e-06,
      "loss": 0.8428,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 369
    },
    {
      "epoch": 0.9319899244332494,
      "grad_norm": 1.0546875,
      "learning_rate": 8.413454386348721e-06,
      "loss": 0.7803,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 370
    },
    {
      "epoch": 0.9345088161209067,
      "grad_norm": 1.0234375,
      "learning_rate": 8.41052170100833e-06,
      "loss": 0.7412,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 371
    },
    {
      "epoch": 0.9370277078085643,
      "grad_norm": 1.046875,
      "learning_rate": 8.407577421480343e-06,
      "loss": 0.9385,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 372
    },
    {
      "epoch": 0.9395465994962217,
      "grad_norm": 1.1328125,
      "learning_rate": 8.40462147887324e-06,
      "loss": 0.7285,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 373
    },
    {
      "epoch": 0.9420654911838791,
      "grad_norm": 1.1015625,
      "learning_rate": 8.401653803748621e-06,
      "loss": 0.7832,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 374
    },
    {
      "epoch": 0.9445843828715366,
      "grad_norm": 1.0703125,
      "learning_rate": 8.398674326115776e-06,
      "loss": 0.7686,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 375
    },
    {
      "epoch": 0.947103274559194,
      "grad_norm": 1.21875,
      "learning_rate": 8.395682975426168e-06,
      "loss": 0.7559,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 376
    },
    {
      "epoch": 0.9496221662468514,
      "grad_norm": 1.1953125,
      "learning_rate": 8.392679680567879e-06,
      "loss": 0.793,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 377
    },
    {
      "epoch": 0.9521410579345088,
      "grad_norm": 1.1328125,
      "learning_rate": 8.389664369859969e-06,
      "loss": 0.7383,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 378
    },
    {
      "epoch": 0.9546599496221663,
      "grad_norm": 5.4375,
      "learning_rate": 8.38663697104677e-06,
      "loss": 0.7324,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 379
    },
    {
      "epoch": 0.9571788413098237,
      "grad_norm": 1.140625,
      "learning_rate": 8.383597411292124e-06,
      "loss": 0.7949,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 380
    },
    {
      "epoch": 0.9596977329974811,
      "grad_norm": 1.109375,
      "learning_rate": 8.380545617173523e-06,
      "loss": 0.7471,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 381
    },
    {
      "epoch": 0.9622166246851386,
      "grad_norm": 1.15625,
      "learning_rate": 8.377481514676227e-06,
      "loss": 0.8252,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 382
    },
    {
      "epoch": 0.964735516372796,
      "grad_norm": 1.0859375,
      "learning_rate": 8.374405029187249e-06,
      "loss": 0.7275,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 383
    },
    {
      "epoch": 0.9672544080604534,
      "grad_norm": 1.3515625,
      "learning_rate": 8.371316085489314e-06,
      "loss": 0.8535,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 384
    },
    {
      "epoch": 0.9697732997481109,
      "grad_norm": 1.0078125,
      "learning_rate": 8.368214607754734e-06,
      "loss": 0.7021,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 385
    },
    {
      "epoch": 0.9722921914357683,
      "grad_norm": 1.046875,
      "learning_rate": 8.365100519539192e-06,
      "loss": 0.7461,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 386
    },
    {
      "epoch": 0.9748110831234257,
      "grad_norm": 1.15625,
      "learning_rate": 8.361973743775462e-06,
      "loss": 0.7598,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 387
    },
    {
      "epoch": 0.9773299748110831,
      "grad_norm": 1.4609375,
      "learning_rate": 8.358834202767068e-06,
      "loss": 0.7588,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 388
    },
    {
      "epoch": 0.9798488664987406,
      "grad_norm": 1.0703125,
      "learning_rate": 8.355681818181818e-06,
      "loss": 0.7598,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 389
    },
    {
      "epoch": 0.982367758186398,
      "grad_norm": 1.203125,
      "learning_rate": 8.35251651104532e-06,
      "loss": 0.7852,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 390
    },
    {
      "epoch": 0.9848866498740554,
      "grad_norm": 1.296875,
      "learning_rate": 8.349338201734368e-06,
      "loss": 0.7285,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 391
    },
    {
      "epoch": 0.9874055415617129,
      "grad_norm": 1.0234375,
      "learning_rate": 8.346146809970272e-06,
      "loss": 0.8027,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 392
    },
    {
      "epoch": 0.9899244332493703,
      "grad_norm": 1.125,
      "learning_rate": 8.342942254812099e-06,
      "loss": 0.749,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 393
    },
    {
      "epoch": 0.9924433249370277,
      "grad_norm": 1.109375,
      "learning_rate": 8.339724454649827e-06,
      "loss": 0.7881,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 394
    },
    {
      "epoch": 0.9949622166246851,
      "grad_norm": 1.71875,
      "learning_rate": 8.336493327197423e-06,
      "loss": 0.7803,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 395
    },
    {
      "epoch": 0.9974811083123426,
      "grad_norm": 1.1953125,
      "learning_rate": 8.33324878948582e-06,
      "loss": 0.8389,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 396
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.203125,
      "learning_rate": 8.329990757855822e-06,
      "loss": 0.8027,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 397
    },
    {
      "epoch": 1.0025188916876575,
      "grad_norm": 1.21875,
      "learning_rate": 8.326719147950914e-06,
      "loss": 0.7734,
      "memory/device_mem_reserved(gib)": 101.22,
      "memory/max_mem_active(gib)": 87.86,
      "memory/max_mem_allocated(gib)": 87.86,
      "step": 398
    }
  ],
  "logging_steps": 1,
  "max_steps": 794,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 199,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6.695614396091073e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}