{
  "best_global_step": 6250,
  "best_metric": 0.98,
  "best_model_checkpoint": "dinov2-Base-finetuned-food101/checkpoint-6250",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 6250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.016,
      "grad_norm": 89.06837463378906,
      "learning_rate": 2.8800000000000004e-07,
      "loss": 5.2089,
      "step": 10
    },
    {
      "epoch": 0.032,
      "grad_norm": 106.91532135009766,
      "learning_rate": 6.08e-07,
      "loss": 5.2135,
      "step": 20
    },
    {
      "epoch": 0.048,
      "grad_norm": 99.38673400878906,
      "learning_rate": 9.28e-07,
      "loss": 4.6396,
      "step": 30
    },
    {
      "epoch": 0.064,
      "grad_norm": 85.49656677246094,
      "learning_rate": 1.248e-06,
      "loss": 4.2156,
      "step": 40
    },
    {
      "epoch": 0.08,
      "grad_norm": 89.49871826171875,
      "learning_rate": 1.568e-06,
      "loss": 3.3982,
      "step": 50
    },
    {
      "epoch": 0.096,
      "grad_norm": 89.63510131835938,
      "learning_rate": 1.8880000000000002e-06,
      "loss": 2.6911,
      "step": 60
    },
    {
      "epoch": 0.112,
      "grad_norm": 83.21440887451172,
      "learning_rate": 2.2080000000000003e-06,
      "loss": 2.1527,
      "step": 70
    },
    {
      "epoch": 0.128,
      "grad_norm": 83.103515625,
      "learning_rate": 2.5280000000000006e-06,
      "loss": 1.5765,
      "step": 80
    },
    {
      "epoch": 0.144,
      "grad_norm": 87.09320831298828,
      "learning_rate": 2.848e-06,
      "loss": 1.255,
      "step": 90
    },
    {
      "epoch": 0.16,
      "grad_norm": 88.66500091552734,
      "learning_rate": 3.1680000000000004e-06,
      "loss": 1.0181,
      "step": 100
    },
    {
      "epoch": 0.176,
      "grad_norm": 112.97905731201172,
      "learning_rate": 3.4880000000000003e-06,
      "loss": 0.8366,
      "step": 110
    },
    {
      "epoch": 0.192,
      "grad_norm": 64.27356719970703,
      "learning_rate": 3.8080000000000006e-06,
      "loss": 0.7367,
      "step": 120
    },
    {
      "epoch": 0.208,
      "grad_norm": 84.11141967773438,
      "learning_rate": 4.128e-06,
      "loss": 0.6918,
      "step": 130
    },
    {
      "epoch": 0.224,
      "grad_norm": 58.329368591308594,
      "learning_rate": 4.4480000000000004e-06,
      "loss": 0.6234,
      "step": 140
    },
    {
      "epoch": 0.24,
      "grad_norm": 81.27049255371094,
      "learning_rate": 4.768000000000001e-06,
      "loss": 0.4478,
      "step": 150
    },
    {
      "epoch": 0.256,
      "grad_norm": 55.579349517822266,
      "learning_rate": 5.088000000000001e-06,
      "loss": 0.5044,
      "step": 160
    },
    {
      "epoch": 0.272,
      "grad_norm": 59.40501403808594,
      "learning_rate": 5.408e-06,
      "loss": 0.4686,
      "step": 170
    },
    {
      "epoch": 0.288,
      "grad_norm": 133.0911407470703,
      "learning_rate": 5.728e-06,
      "loss": 0.3406,
      "step": 180
    },
    {
      "epoch": 0.304,
      "grad_norm": 51.79353332519531,
      "learning_rate": 6.048e-06,
      "loss": 0.5976,
      "step": 190
    },
    {
      "epoch": 0.32,
      "grad_norm": 80.84319305419922,
      "learning_rate": 6.368000000000001e-06,
      "loss": 0.5719,
      "step": 200
    },
    {
      "epoch": 0.336,
      "grad_norm": 46.885093688964844,
      "learning_rate": 6.688e-06,
      "loss": 0.3955,
      "step": 210
    },
    {
      "epoch": 0.352,
      "grad_norm": 74.30220794677734,
      "learning_rate": 7.0080000000000005e-06,
      "loss": 0.4505,
      "step": 220
    },
    {
      "epoch": 0.368,
      "grad_norm": 85.305908203125,
      "learning_rate": 7.328000000000001e-06,
      "loss": 0.3513,
      "step": 230
    },
    {
      "epoch": 0.384,
      "grad_norm": 57.64317321777344,
      "learning_rate": 7.648e-06,
      "loss": 0.4213,
      "step": 240
    },
    {
      "epoch": 0.4,
      "grad_norm": 32.956275939941406,
      "learning_rate": 7.968e-06,
      "loss": 0.3075,
      "step": 250
    },
    {
      "epoch": 0.416,
      "grad_norm": 68.65741729736328,
      "learning_rate": 8.288000000000001e-06,
      "loss": 0.3231,
      "step": 260
    },
    {
      "epoch": 0.432,
      "grad_norm": 61.947906494140625,
      "learning_rate": 8.608000000000001e-06,
      "loss": 0.3555,
      "step": 270
    },
    {
      "epoch": 0.448,
      "grad_norm": 49.08671951293945,
      "learning_rate": 8.928000000000002e-06,
      "loss": 0.529,
      "step": 280
    },
    {
      "epoch": 0.464,
      "grad_norm": 88.26673889160156,
      "learning_rate": 9.248e-06,
      "loss": 0.4929,
      "step": 290
    },
    {
      "epoch": 0.48,
      "grad_norm": 82.20326232910156,
      "learning_rate": 9.568e-06,
      "loss": 0.6569,
      "step": 300
    },
    {
      "epoch": 0.496,
      "grad_norm": 22.57343292236328,
      "learning_rate": 9.888000000000001e-06,
      "loss": 0.3489,
      "step": 310
    },
    {
      "epoch": 0.512,
      "grad_norm": 61.01542282104492,
      "learning_rate": 1.0208e-05,
      "loss": 0.6376,
      "step": 320
    },
    {
      "epoch": 0.528,
      "grad_norm": 11.429601669311523,
      "learning_rate": 1.0528e-05,
      "loss": 0.4121,
      "step": 330
    },
    {
      "epoch": 0.544,
      "grad_norm": 43.226837158203125,
      "learning_rate": 1.0848e-05,
      "loss": 0.4455,
      "step": 340
    },
    {
      "epoch": 0.56,
      "grad_norm": 101.21278381347656,
      "learning_rate": 1.1168e-05,
      "loss": 0.4205,
      "step": 350
    },
    {
      "epoch": 0.576,
      "grad_norm": 45.80269241333008,
      "learning_rate": 1.1488e-05,
      "loss": 0.4058,
      "step": 360
    },
    {
      "epoch": 0.592,
      "grad_norm": 88.07817077636719,
      "learning_rate": 1.1808000000000001e-05,
      "loss": 0.516,
      "step": 370
    },
    {
      "epoch": 0.608,
      "grad_norm": 117.33232116699219,
      "learning_rate": 1.2128000000000001e-05,
      "loss": 0.4378,
      "step": 380
    },
    {
      "epoch": 0.624,
      "grad_norm": 193.94338989257812,
      "learning_rate": 1.2448e-05,
      "loss": 0.5074,
      "step": 390
    },
    {
      "epoch": 0.64,
      "grad_norm": 48.7374267578125,
      "learning_rate": 1.2768e-05,
      "loss": 0.3549,
      "step": 400
    },
    {
      "epoch": 0.656,
      "grad_norm": 11.20517635345459,
      "learning_rate": 1.3088e-05,
      "loss": 0.3929,
      "step": 410
    },
    {
      "epoch": 0.672,
      "grad_norm": 43.28117370605469,
      "learning_rate": 1.3408000000000001e-05,
      "loss": 0.2347,
      "step": 420
    },
    {
      "epoch": 0.688,
      "grad_norm": 54.67144012451172,
      "learning_rate": 1.3728000000000001e-05,
      "loss": 0.4916,
      "step": 430
    },
    {
      "epoch": 0.704,
      "grad_norm": 177.31204223632812,
      "learning_rate": 1.4048000000000002e-05,
      "loss": 0.5148,
      "step": 440
    },
    {
      "epoch": 0.72,
      "grad_norm": 57.29036331176758,
      "learning_rate": 1.4368000000000002e-05,
      "loss": 0.4149,
      "step": 450
    },
    {
      "epoch": 0.736,
      "grad_norm": 71.80987548828125,
      "learning_rate": 1.4688000000000002e-05,
      "loss": 0.3869,
      "step": 460
    },
    {
      "epoch": 0.752,
      "grad_norm": 35.79969024658203,
      "learning_rate": 1.5008000000000001e-05,
      "loss": 0.4098,
      "step": 470
    },
    {
      "epoch": 0.768,
      "grad_norm": 76.79283905029297,
      "learning_rate": 1.5328e-05,
      "loss": 0.3789,
      "step": 480
    },
    {
      "epoch": 0.784,
      "grad_norm": 4.4785943031311035,
      "learning_rate": 1.5648e-05,
      "loss": 0.3875,
      "step": 490
    },
    {
      "epoch": 0.8,
      "grad_norm": 222.21080017089844,
      "learning_rate": 1.5968e-05,
      "loss": 0.5796,
      "step": 500
    },
    {
      "epoch": 0.816,
      "grad_norm": 85.26457214355469,
      "learning_rate": 1.6288e-05,
      "loss": 0.4111,
      "step": 510
    },
    {
      "epoch": 0.832,
      "grad_norm": 19.47867202758789,
      "learning_rate": 1.6608e-05,
      "loss": 0.4095,
      "step": 520
    },
    {
      "epoch": 0.848,
      "grad_norm": 91.59925842285156,
      "learning_rate": 1.6928e-05,
      "loss": 0.5244,
      "step": 530
    },
    {
      "epoch": 0.864,
      "grad_norm": 20.435232162475586,
      "learning_rate": 1.7248e-05,
      "loss": 0.4291,
      "step": 540
    },
    {
      "epoch": 0.88,
      "grad_norm": 104.42509460449219,
      "learning_rate": 1.7568000000000002e-05,
      "loss": 0.5047,
      "step": 550
    },
    {
      "epoch": 0.896,
      "grad_norm": 142.441162109375,
      "learning_rate": 1.7888000000000002e-05,
      "loss": 0.5301,
      "step": 560
    },
    {
      "epoch": 0.912,
      "grad_norm": 95.12364196777344,
      "learning_rate": 1.8208000000000003e-05,
      "loss": 0.513,
      "step": 570
    },
    {
      "epoch": 0.928,
      "grad_norm": 98.76264953613281,
      "learning_rate": 1.8528000000000003e-05,
      "loss": 0.6661,
      "step": 580
    },
    {
      "epoch": 0.944,
      "grad_norm": 65.93214416503906,
      "learning_rate": 1.8848000000000003e-05,
      "loss": 0.6454,
      "step": 590
    },
    {
      "epoch": 0.96,
      "grad_norm": 178.0699462890625,
      "learning_rate": 1.9168000000000004e-05,
      "loss": 0.375,
      "step": 600
    },
    {
      "epoch": 0.976,
      "grad_norm": 82.4294204711914,
      "learning_rate": 1.9488000000000004e-05,
      "loss": 0.578,
      "step": 610
    },
    {
      "epoch": 0.992,
      "grad_norm": 68.52827453613281,
      "learning_rate": 1.9808e-05,
      "loss": 0.675,
      "step": 620
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.938,
      "eval_f1": 0.9398621531246216,
      "eval_loss": 0.17777465283870697,
      "eval_runtime": 25.103,
      "eval_samples_per_second": 59.754,
      "eval_steps_per_second": 14.938,
      "step": 625
    },
    {
      "epoch": 1.008,
      "grad_norm": 90.21050262451172,
      "learning_rate": 1.998577777777778e-05,
      "loss": 0.5093,
      "step": 630
    },
    {
      "epoch": 1.024,
      "grad_norm": 68.47109985351562,
      "learning_rate": 1.9950222222222225e-05,
      "loss": 0.4518,
      "step": 640
    },
    {
      "epoch": 1.04,
      "grad_norm": 65.01594543457031,
      "learning_rate": 1.9914666666666668e-05,
      "loss": 0.5236,
      "step": 650
    },
    {
      "epoch": 1.056,
      "grad_norm": 91.11334991455078,
      "learning_rate": 1.9879111111111113e-05,
      "loss": 0.5741,
      "step": 660
    },
    {
      "epoch": 1.072,
      "grad_norm": 69.707763671875,
      "learning_rate": 1.984355555555556e-05,
      "loss": 0.5296,
      "step": 670
    },
    {
      "epoch": 1.088,
      "grad_norm": 82.90894317626953,
      "learning_rate": 1.9808e-05,
      "loss": 0.6058,
      "step": 680
    },
    {
      "epoch": 1.104,
      "grad_norm": 139.13919067382812,
      "learning_rate": 1.9772444444444446e-05,
      "loss": 0.6164,
      "step": 690
    },
    {
      "epoch": 1.12,
      "grad_norm": 30.786418914794922,
      "learning_rate": 1.973688888888889e-05,
      "loss": 0.39,
      "step": 700
    },
    {
      "epoch": 1.1360000000000001,
      "grad_norm": 73.54574584960938,
      "learning_rate": 1.9701333333333334e-05,
      "loss": 0.4312,
      "step": 710
    },
    {
      "epoch": 1.152,
      "grad_norm": 43.80199432373047,
      "learning_rate": 1.966577777777778e-05,
      "loss": 0.375,
      "step": 720
    },
    {
      "epoch": 1.168,
      "grad_norm": 152.24276733398438,
      "learning_rate": 1.9630222222222225e-05,
      "loss": 0.2291,
      "step": 730
    },
    {
      "epoch": 1.184,
      "grad_norm": 73.12702941894531,
      "learning_rate": 1.9594666666666667e-05,
      "loss": 0.3965,
      "step": 740
    },
    {
      "epoch": 1.2,
      "grad_norm": 61.434120178222656,
      "learning_rate": 1.9559111111111113e-05,
      "loss": 0.4949,
      "step": 750
    },
    {
      "epoch": 1.216,
      "grad_norm": 57.76673126220703,
      "learning_rate": 1.9523555555555558e-05,
      "loss": 0.537,
      "step": 760
    },
    {
      "epoch": 1.232,
      "grad_norm": 63.07472229003906,
      "learning_rate": 1.9488000000000004e-05,
      "loss": 0.3069,
      "step": 770
    },
    {
      "epoch": 1.248,
      "grad_norm": 82.42784118652344,
      "learning_rate": 1.9452444444444446e-05,
      "loss": 0.4934,
      "step": 780
    },
    {
      "epoch": 1.264,
      "grad_norm": 109.37876892089844,
      "learning_rate": 1.9416888888888888e-05,
      "loss": 0.4084,
      "step": 790
    },
    {
      "epoch": 1.28,
      "grad_norm": 38.176002502441406,
      "learning_rate": 1.9381333333333334e-05,
      "loss": 0.4815,
      "step": 800
    },
    {
      "epoch": 1.296,
      "grad_norm": 83.43049621582031,
      "learning_rate": 1.934577777777778e-05,
      "loss": 0.5391,
      "step": 810
    },
    {
      "epoch": 1.312,
      "grad_norm": 1.9186406135559082,
      "learning_rate": 1.9310222222222225e-05,
      "loss": 0.3134,
      "step": 820
    },
    {
      "epoch": 1.328,
      "grad_norm": 81.61128234863281,
      "learning_rate": 1.9274666666666667e-05,
      "loss": 0.4807,
      "step": 830
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 37.480411529541016,
      "learning_rate": 1.9239111111111112e-05,
      "loss": 0.448,
      "step": 840
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 80.07466125488281,
      "learning_rate": 1.9203555555555558e-05,
      "loss": 0.3658,
      "step": 850
    },
    {
      "epoch": 1.376,
      "grad_norm": 55.322166442871094,
      "learning_rate": 1.9168000000000004e-05,
      "loss": 0.3584,
      "step": 860
    },
    {
      "epoch": 1.392,
      "grad_norm": 2.566350221633911,
      "learning_rate": 1.9132444444444446e-05,
      "loss": 0.2511,
      "step": 870
    },
    {
      "epoch": 1.408,
      "grad_norm": 25.400840759277344,
      "learning_rate": 1.909688888888889e-05,
      "loss": 0.2569,
      "step": 880
    },
    {
      "epoch": 1.424,
      "grad_norm": 21.915630340576172,
      "learning_rate": 1.9061333333333333e-05,
      "loss": 0.3917,
      "step": 890
    },
    {
      "epoch": 1.44,
      "grad_norm": 80.36392211914062,
      "learning_rate": 1.902577777777778e-05,
      "loss": 0.307,
      "step": 900
    },
    {
      "epoch": 1.456,
      "grad_norm": 167.50462341308594,
      "learning_rate": 1.8990222222222224e-05,
      "loss": 0.4593,
      "step": 910
    },
    {
      "epoch": 1.472,
      "grad_norm": 108.12252807617188,
      "learning_rate": 1.8954666666666667e-05,
      "loss": 0.5507,
      "step": 920
    },
    {
      "epoch": 1.488,
      "grad_norm": 43.201087951660156,
      "learning_rate": 1.8919111111111112e-05,
      "loss": 0.4131,
      "step": 930
    },
    {
      "epoch": 1.504,
      "grad_norm": 66.8707046508789,
      "learning_rate": 1.8883555555555558e-05,
      "loss": 0.3105,
      "step": 940
    },
    {
      "epoch": 1.52,
      "grad_norm": 23.931177139282227,
      "learning_rate": 1.8848000000000003e-05,
      "loss": 0.374,
      "step": 950
    },
    {
      "epoch": 1.536,
      "grad_norm": 112.31582641601562,
      "learning_rate": 1.8812444444444445e-05,
      "loss": 0.4161,
      "step": 960
    },
    {
      "epoch": 1.552,
      "grad_norm": 88.35121154785156,
      "learning_rate": 1.877688888888889e-05,
      "loss": 0.6474,
      "step": 970
    },
    {
      "epoch": 1.568,
      "grad_norm": 76.46733093261719,
      "learning_rate": 1.8741333333333336e-05,
      "loss": 0.3997,
      "step": 980
    },
    {
      "epoch": 1.584,
      "grad_norm": 110.73181915283203,
      "learning_rate": 1.870577777777778e-05,
      "loss": 0.6053,
      "step": 990
    },
    {
      "epoch": 1.6,
      "grad_norm": 35.111751556396484,
      "learning_rate": 1.8670222222222224e-05,
      "loss": 0.2589,
      "step": 1000
    },
    {
      "epoch": 1.616,
      "grad_norm": 77.90402221679688,
      "learning_rate": 1.8634666666666666e-05,
      "loss": 0.485,
      "step": 1010
    },
    {
      "epoch": 1.6320000000000001,
      "grad_norm": 5.89962100982666,
      "learning_rate": 1.8599111111111112e-05,
      "loss": 0.4286,
      "step": 1020
    },
    {
      "epoch": 1.6480000000000001,
      "grad_norm": 50.08778762817383,
      "learning_rate": 1.8563555555555557e-05,
      "loss": 0.3526,
      "step": 1030
    },
    {
      "epoch": 1.6640000000000001,
      "grad_norm": 36.45596694946289,
      "learning_rate": 1.8528000000000003e-05,
      "loss": 0.38,
      "step": 1040
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 33.204345703125,
      "learning_rate": 1.8492444444444445e-05,
      "loss": 0.4526,
      "step": 1050
    },
    {
      "epoch": 1.696,
      "grad_norm": 49.752349853515625,
      "learning_rate": 1.845688888888889e-05,
      "loss": 0.3143,
      "step": 1060
    },
    {
      "epoch": 1.712,
      "grad_norm": 69.24263763427734,
      "learning_rate": 1.8421333333333336e-05,
      "loss": 0.3269,
      "step": 1070
    },
    {
      "epoch": 1.728,
      "grad_norm": 191.12741088867188,
      "learning_rate": 1.838577777777778e-05,
      "loss": 0.4755,
      "step": 1080
    },
    {
      "epoch": 1.744,
      "grad_norm": 86.41985321044922,
      "learning_rate": 1.8350222222222224e-05,
      "loss": 0.6529,
      "step": 1090
    },
    {
      "epoch": 1.76,
      "grad_norm": 102.5144271850586,
      "learning_rate": 1.8314666666666666e-05,
      "loss": 0.4386,
      "step": 1100
    },
    {
      "epoch": 1.776,
      "grad_norm": 124.9957275390625,
      "learning_rate": 1.827911111111111e-05,
      "loss": 0.5639,
      "step": 1110
    },
    {
      "epoch": 1.792,
      "grad_norm": 21.024431228637695,
      "learning_rate": 1.8243555555555557e-05,
      "loss": 0.3691,
      "step": 1120
    },
    {
      "epoch": 1.808,
      "grad_norm": 68.7905502319336,
      "learning_rate": 1.8208000000000003e-05,
      "loss": 0.4668,
      "step": 1130
    },
    {
      "epoch": 1.8239999999999998,
      "grad_norm": 58.76439666748047,
      "learning_rate": 1.8172444444444445e-05,
      "loss": 0.5254,
      "step": 1140
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 52.31269073486328,
      "learning_rate": 1.813688888888889e-05,
      "loss": 0.3713,
      "step": 1150
    },
    {
      "epoch": 1.8559999999999999,
      "grad_norm": 47.486427307128906,
      "learning_rate": 1.8101333333333336e-05,
      "loss": 0.4245,
      "step": 1160
    },
    {
      "epoch": 1.8719999999999999,
      "grad_norm": 68.47742462158203,
      "learning_rate": 1.806577777777778e-05,
      "loss": 0.6182,
      "step": 1170
    },
    {
      "epoch": 1.888,
      "grad_norm": 65.69060516357422,
      "learning_rate": 1.8030222222222223e-05,
      "loss": 0.499,
      "step": 1180
    },
    {
      "epoch": 1.904,
      "grad_norm": 76.0249252319336,
      "learning_rate": 1.799466666666667e-05,
      "loss": 0.3416,
      "step": 1190
    },
    {
      "epoch": 1.92,
      "grad_norm": 46.81635284423828,
      "learning_rate": 1.795911111111111e-05,
      "loss": 0.413,
      "step": 1200
    },
    {
      "epoch": 1.936,
      "grad_norm": 45.73046112060547,
      "learning_rate": 1.7923555555555557e-05,
      "loss": 0.4074,
      "step": 1210
    },
    {
      "epoch": 1.952,
      "grad_norm": 35.92128372192383,
      "learning_rate": 1.7888000000000002e-05,
      "loss": 0.428,
      "step": 1220
    },
    {
      "epoch": 1.968,
      "grad_norm": 54.154273986816406,
      "learning_rate": 1.7852444444444444e-05,
      "loss": 0.3654,
      "step": 1230
    },
    {
      "epoch": 1.984,
      "grad_norm": 117.35140228271484,
      "learning_rate": 1.781688888888889e-05,
      "loss": 0.5069,
      "step": 1240
    },
    {
      "epoch": 2.0,
      "grad_norm": 69.95264434814453,
      "learning_rate": 1.7781333333333335e-05,
      "loss": 0.5284,
      "step": 1250
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9613333333333334,
      "eval_f1": 0.9614742314125441,
      "eval_loss": 0.12478982657194138,
      "eval_runtime": 24.9506,
      "eval_samples_per_second": 60.119,
      "eval_steps_per_second": 15.03,
      "step": 1250
    },
    {
      "epoch": 2.016,
      "grad_norm": 90.76451873779297,
      "learning_rate": 1.774577777777778e-05,
      "loss": 0.4187,
      "step": 1260
    },
    {
      "epoch": 2.032,
      "grad_norm": 22.007369995117188,
      "learning_rate": 1.7710222222222223e-05,
      "loss": 0.2309,
      "step": 1270
    },
    {
      "epoch": 2.048,
      "grad_norm": 56.51552963256836,
      "learning_rate": 1.767466666666667e-05,
      "loss": 0.3627,
      "step": 1280
    },
    {
      "epoch": 2.064,
      "grad_norm": 52.548004150390625,
      "learning_rate": 1.7639111111111114e-05,
      "loss": 0.4389,
      "step": 1290
    },
    {
      "epoch": 2.08,
      "grad_norm": 54.10152053833008,
      "learning_rate": 1.7603555555555556e-05,
      "loss": 0.4167,
      "step": 1300
    },
    {
      "epoch": 2.096,
      "grad_norm": 55.00556564331055,
      "learning_rate": 1.7568000000000002e-05,
      "loss": 0.3528,
      "step": 1310
    },
    {
      "epoch": 2.112,
      "grad_norm": 12.216611862182617,
      "learning_rate": 1.7532444444444444e-05,
      "loss": 0.1805,
      "step": 1320
    },
    {
      "epoch": 2.128,
      "grad_norm": 114.3968505859375,
      "learning_rate": 1.749688888888889e-05,
      "loss": 0.3243,
      "step": 1330
    },
    {
      "epoch": 2.144,
      "grad_norm": 5.575631141662598,
      "learning_rate": 1.7461333333333335e-05,
      "loss": 0.4095,
      "step": 1340
    },
    {
      "epoch": 2.16,
      "grad_norm": 97.60520935058594,
      "learning_rate": 1.742577777777778e-05,
      "loss": 0.4591,
      "step": 1350
    },
    {
      "epoch": 2.176,
      "grad_norm": 35.92754364013672,
      "learning_rate": 1.7390222222222223e-05,
      "loss": 0.3305,
      "step": 1360
    },
    {
      "epoch": 2.192,
      "grad_norm": 46.582794189453125,
      "learning_rate": 1.735466666666667e-05,
      "loss": 0.4082,
      "step": 1370
    },
    {
      "epoch": 2.208,
      "grad_norm": 73.6832504272461,
      "learning_rate": 1.7319111111111114e-05,
      "loss": 0.3443,
      "step": 1380
    },
    {
      "epoch": 2.224,
      "grad_norm": 110.241455078125,
      "learning_rate": 1.728355555555556e-05,
      "loss": 0.2277,
      "step": 1390
    },
    {
      "epoch": 2.24,
      "grad_norm": 30.731782913208008,
      "learning_rate": 1.7248e-05,
      "loss": 0.2708,
      "step": 1400
    },
    {
      "epoch": 2.2560000000000002,
      "grad_norm": 15.310711860656738,
      "learning_rate": 1.7212444444444444e-05,
      "loss": 0.3453,
      "step": 1410
    },
    {
      "epoch": 2.2720000000000002,
      "grad_norm": 50.28248977661133,
      "learning_rate": 1.717688888888889e-05,
      "loss": 0.2364,
      "step": 1420
    },
    {
      "epoch": 2.288,
      "grad_norm": 57.56525802612305,
      "learning_rate": 1.7141333333333335e-05,
      "loss": 0.2759,
      "step": 1430
    },
    {
      "epoch": 2.304,
      "grad_norm": 63.90391540527344,
      "learning_rate": 1.710577777777778e-05,
      "loss": 0.3155,
      "step": 1440
    },
    {
      "epoch": 2.32,
      "grad_norm": 28.447397232055664,
      "learning_rate": 1.7070222222222222e-05,
      "loss": 0.4112,
      "step": 1450
    },
    {
      "epoch": 2.336,
      "grad_norm": 51.297603607177734,
      "learning_rate": 1.7034666666666668e-05,
      "loss": 0.29,
      "step": 1460
    },
    {
      "epoch": 2.352,
      "grad_norm": 29.22579002380371,
      "learning_rate": 1.6999111111111114e-05,
      "loss": 0.3201,
      "step": 1470
    },
    {
      "epoch": 2.368,
      "grad_norm": 49.331199645996094,
      "learning_rate": 1.696355555555556e-05,
      "loss": 0.4024,
      "step": 1480
    },
    {
      "epoch": 2.384,
      "grad_norm": 45.29281234741211,
      "learning_rate": 1.6928e-05,
      "loss": 0.2184,
      "step": 1490
    },
    {
      "epoch": 2.4,
      "grad_norm": 63.356895446777344,
      "learning_rate": 1.6892444444444447e-05,
      "loss": 0.2487,
      "step": 1500
    },
    {
      "epoch": 2.416,
      "grad_norm": 3.2891721725463867,
      "learning_rate": 1.685688888888889e-05,
      "loss": 0.2616,
      "step": 1510
    },
    {
      "epoch": 2.432,
      "grad_norm": 66.31077575683594,
      "learning_rate": 1.6821333333333334e-05,
      "loss": 0.3052,
      "step": 1520
    },
    {
      "epoch": 2.448,
      "grad_norm": 105.19303894042969,
      "learning_rate": 1.678577777777778e-05,
      "loss": 0.4421,
      "step": 1530
    },
    {
      "epoch": 2.464,
      "grad_norm": 31.327606201171875,
      "learning_rate": 1.6750222222222222e-05,
      "loss": 0.3093,
      "step": 1540
    },
    {
      "epoch": 2.48,
      "grad_norm": 78.41735076904297,
      "learning_rate": 1.6714666666666668e-05,
      "loss": 0.5758,
      "step": 1550
    },
    {
      "epoch": 2.496,
      "grad_norm": 45.926918029785156,
      "learning_rate": 1.6679111111111113e-05,
      "loss": 0.2822,
      "step": 1560
    },
    {
      "epoch": 2.512,
      "grad_norm": 90.95999908447266,
      "learning_rate": 1.664355555555556e-05,
      "loss": 0.3204,
      "step": 1570
    },
    {
      "epoch": 2.528,
      "grad_norm": 16.62053871154785,
      "learning_rate": 1.6608e-05,
      "loss": 0.3205,
      "step": 1580
    },
    {
      "epoch": 2.544,
      "grad_norm": 163.37112426757812,
      "learning_rate": 1.6572444444444446e-05,
      "loss": 0.2595,
      "step": 1590
    },
    {
      "epoch": 2.56,
      "grad_norm": 5.807415008544922,
      "learning_rate": 1.6536888888888892e-05,
      "loss": 0.2315,
      "step": 1600
    },
    {
      "epoch": 2.576,
      "grad_norm": 32.90766525268555,
      "learning_rate": 1.6501333333333334e-05,
      "loss": 0.3351,
      "step": 1610
    },
    {
      "epoch": 2.592,
      "grad_norm": 53.188724517822266,
      "learning_rate": 1.646577777777778e-05,
      "loss": 0.2578,
      "step": 1620
    },
    {
      "epoch": 2.608,
      "grad_norm": 29.000858306884766,
      "learning_rate": 1.6430222222222222e-05,
      "loss": 0.4,
      "step": 1630
    },
    {
      "epoch": 2.624,
      "grad_norm": 7.635106563568115,
      "learning_rate": 1.6394666666666667e-05,
      "loss": 0.315,
      "step": 1640
    },
    {
      "epoch": 2.64,
      "grad_norm": 2.9866526126861572,
      "learning_rate": 1.6359111111111113e-05,
      "loss": 0.2032,
      "step": 1650
    },
    {
      "epoch": 2.656,
      "grad_norm": 68.7740707397461,
      "learning_rate": 1.632355555555556e-05,
      "loss": 0.508,
      "step": 1660
    },
    {
      "epoch": 2.672,
      "grad_norm": 58.381858825683594,
      "learning_rate": 1.6288e-05,
      "loss": 0.4532,
      "step": 1670
    },
    {
      "epoch": 2.6879999999999997,
      "grad_norm": 83.0326156616211,
      "learning_rate": 1.6252444444444446e-05,
      "loss": 0.3074,
      "step": 1680
    },
    {
      "epoch": 2.7039999999999997,
      "grad_norm": 48.644752502441406,
      "learning_rate": 1.621688888888889e-05,
      "loss": 0.2919,
      "step": 1690
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 43.27705001831055,
      "learning_rate": 1.6181333333333337e-05,
      "loss": 0.4442,
      "step": 1700
    },
    {
      "epoch": 2.7359999999999998,
      "grad_norm": 62.09077072143555,
      "learning_rate": 1.614577777777778e-05,
      "loss": 0.3772,
      "step": 1710
    },
    {
      "epoch": 2.752,
      "grad_norm": 97.17964935302734,
      "learning_rate": 1.611022222222222e-05,
      "loss": 0.2832,
      "step": 1720
    },
    {
      "epoch": 2.768,
      "grad_norm": 113.52243041992188,
      "learning_rate": 1.6074666666666667e-05,
      "loss": 0.3146,
      "step": 1730
    },
    {
      "epoch": 2.784,
      "grad_norm": 46.1737060546875,
      "learning_rate": 1.6039111111111113e-05,
      "loss": 0.6467,
      "step": 1740
    },
    {
      "epoch": 2.8,
      "grad_norm": 53.177696228027344,
      "learning_rate": 1.6003555555555558e-05,
      "loss": 0.3414,
      "step": 1750
    },
    {
      "epoch": 2.816,
      "grad_norm": 72.3310775756836,
      "learning_rate": 1.5968e-05,
      "loss": 0.3393,
      "step": 1760
    },
    {
      "epoch": 2.832,
      "grad_norm": 24.155517578125,
      "learning_rate": 1.5932444444444446e-05,
      "loss": 0.2632,
      "step": 1770
    },
    {
      "epoch": 2.848,
      "grad_norm": 71.61637878417969,
      "learning_rate": 1.589688888888889e-05,
      "loss": 0.2974,
      "step": 1780
    },
    {
      "epoch": 2.864,
      "grad_norm": 47.56396484375,
      "learning_rate": 1.5861333333333337e-05,
      "loss": 0.3304,
      "step": 1790
    },
    {
      "epoch": 2.88,
      "grad_norm": 52.59245300292969,
      "learning_rate": 1.582577777777778e-05,
      "loss": 0.2938,
      "step": 1800
    },
    {
      "epoch": 2.896,
      "grad_norm": 68.7037582397461,
      "learning_rate": 1.5790222222222225e-05,
      "loss": 0.2522,
      "step": 1810
    },
    {
      "epoch": 2.912,
      "grad_norm": 62.64757537841797,
      "learning_rate": 1.5754666666666667e-05,
      "loss": 0.2989,
      "step": 1820
    },
    {
      "epoch": 2.928,
      "grad_norm": 58.7310791015625,
      "learning_rate": 1.5719111111111112e-05,
      "loss": 0.175,
      "step": 1830
    },
    {
      "epoch": 2.944,
      "grad_norm": 0.8208155035972595,
      "learning_rate": 1.5683555555555558e-05,
      "loss": 0.454,
      "step": 1840
    },
    {
      "epoch": 2.96,
      "grad_norm": 53.65021514892578,
      "learning_rate": 1.5648e-05,
      "loss": 0.2396,
      "step": 1850
    },
    {
      "epoch": 2.976,
      "grad_norm": 48.882816314697266,
      "learning_rate": 1.5612444444444445e-05,
      "loss": 0.412,
      "step": 1860
    },
    {
      "epoch": 2.992,
      "grad_norm": 75.47936248779297,
      "learning_rate": 1.557688888888889e-05,
      "loss": 0.3455,
      "step": 1870
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9513333333333334,
      "eval_f1": 0.9522551210889584,
      "eval_loss": 0.1548186093568802,
      "eval_runtime": 25.3019,
      "eval_samples_per_second": 59.284,
      "eval_steps_per_second": 14.821,
      "step": 1875
    },
    {
      "epoch": 3.008,
      "grad_norm": 14.862210273742676,
      "learning_rate": 1.5541333333333337e-05,
      "loss": 0.3599,
      "step": 1880
    },
    {
      "epoch": 3.024,
      "grad_norm": 39.82673263549805,
      "learning_rate": 1.550577777777778e-05,
      "loss": 0.4111,
      "step": 1890
    },
    {
      "epoch": 3.04,
      "grad_norm": 64.49237823486328,
      "learning_rate": 1.5470222222222224e-05,
      "loss": 0.1988,
      "step": 1900
    },
    {
      "epoch": 3.056,
      "grad_norm": 53.11403274536133,
      "learning_rate": 1.543466666666667e-05,
      "loss": 0.292,
      "step": 1910
    },
    {
      "epoch": 3.072,
      "grad_norm": 61.9869499206543,
      "learning_rate": 1.5399111111111112e-05,
      "loss": 0.1491,
      "step": 1920
    },
    {
      "epoch": 3.088,
      "grad_norm": 24.670124053955078,
      "learning_rate": 1.5363555555555557e-05,
      "loss": 0.3061,
      "step": 1930
    },
    {
      "epoch": 3.104,
      "grad_norm": 5.021854877471924,
      "learning_rate": 1.5328e-05,
      "loss": 0.3416,
      "step": 1940
    },
    {
      "epoch": 3.12,
      "grad_norm": 53.80380630493164,
      "learning_rate": 1.5292444444444445e-05,
      "loss": 0.2582,
      "step": 1950
    },
    {
      "epoch": 3.136,
      "grad_norm": 43.22938537597656,
      "learning_rate": 1.525688888888889e-05,
      "loss": 0.131,
      "step": 1960
    },
    {
      "epoch": 3.152,
      "grad_norm": 51.80830001831055,
      "learning_rate": 1.5221333333333335e-05,
      "loss": 0.4225,
      "step": 1970
    },
    {
      "epoch": 3.168,
      "grad_norm": 51.97641372680664,
      "learning_rate": 1.518577777777778e-05,
      "loss": 0.2496,
      "step": 1980
    },
    {
      "epoch": 3.184,
      "grad_norm": 78.88411712646484,
      "learning_rate": 1.5150222222222224e-05,
      "loss": 0.2843,
      "step": 1990
    },
    {
      "epoch": 3.2,
      "grad_norm": 80.55473327636719,
      "learning_rate": 1.5114666666666668e-05,
      "loss": 0.2579,
      "step": 2000
    },
    {
      "epoch": 3.216,
      "grad_norm": 10.610305786132812,
      "learning_rate": 1.5079111111111113e-05,
      "loss": 0.251,
      "step": 2010
    },
    {
      "epoch": 3.232,
      "grad_norm": 18.754613876342773,
      "learning_rate": 1.5043555555555555e-05,
      "loss": 0.2215,
      "step": 2020
    },
    {
      "epoch": 3.248,
      "grad_norm": 91.42521667480469,
      "learning_rate": 1.5008000000000001e-05,
      "loss": 0.4424,
      "step": 2030
    },
    {
      "epoch": 3.2640000000000002,
      "grad_norm": 72.18260955810547,
      "learning_rate": 1.4972444444444445e-05,
      "loss": 0.2405,
      "step": 2040
    },
    {
      "epoch": 3.2800000000000002,
      "grad_norm": 25.848726272583008,
      "learning_rate": 1.493688888888889e-05,
      "loss": 0.1684,
      "step": 2050
    },
    {
      "epoch": 3.296,
      "grad_norm": 42.30290222167969,
      "learning_rate": 1.4901333333333334e-05,
      "loss": 0.1324,
      "step": 2060
    },
    {
      "epoch": 3.312,
      "grad_norm": 121.94938659667969,
      "learning_rate": 1.486577777777778e-05,
      "loss": 0.3352,
      "step": 2070
    },
    {
      "epoch": 3.328,
      "grad_norm": 26.369827270507812,
      "learning_rate": 1.4830222222222224e-05,
      "loss": 0.264,
      "step": 2080
    },
    {
      "epoch": 3.344,
      "grad_norm": 77.36036682128906,
      "learning_rate": 1.4794666666666669e-05,
      "loss": 0.4048,
      "step": 2090
    },
    {
      "epoch": 3.36,
      "grad_norm": 68.10829162597656,
      "learning_rate": 1.4759111111111113e-05,
      "loss": 0.255,
      "step": 2100
    },
    {
      "epoch": 3.376,
      "grad_norm": 50.70305633544922,
      "learning_rate": 1.4723555555555557e-05,
      "loss": 0.2573,
      "step": 2110
    },
    {
      "epoch": 3.392,
      "grad_norm": 17.606157302856445,
      "learning_rate": 1.4688000000000002e-05,
      "loss": 0.1742,
      "step": 2120
    },
    {
      "epoch": 3.408,
      "grad_norm": 47.515872955322266,
      "learning_rate": 1.4652444444444445e-05,
      "loss": 0.3382,
      "step": 2130
    },
    {
      "epoch": 3.424,
      "grad_norm": 35.3089599609375,
      "learning_rate": 1.461688888888889e-05,
      "loss": 0.2477,
      "step": 2140
    },
    {
      "epoch": 3.44,
      "grad_norm": 53.55549240112305,
      "learning_rate": 1.4581333333333334e-05,
      "loss": 0.1245,
      "step": 2150
    },
    {
      "epoch": 3.456,
      "grad_norm": 52.0760498046875,
      "learning_rate": 1.454577777777778e-05,
      "loss": 0.2309,
      "step": 2160
    },
    {
      "epoch": 3.472,
      "grad_norm": 12.876019477844238,
      "learning_rate": 1.4510222222222223e-05,
      "loss": 0.2314,
      "step": 2170
    },
    {
      "epoch": 3.488,
      "grad_norm": 126.93236541748047,
      "learning_rate": 1.4474666666666669e-05,
      "loss": 0.363,
      "step": 2180
    },
    {
      "epoch": 3.504,
      "grad_norm": 33.92259979248047,
      "learning_rate": 1.4439111111111113e-05,
      "loss": 0.1426,
      "step": 2190
    },
    {
      "epoch": 3.52,
      "grad_norm": 39.19934844970703,
      "learning_rate": 1.4403555555555556e-05,
      "loss": 0.4135,
      "step": 2200
    },
    {
      "epoch": 3.536,
      "grad_norm": 95.46045684814453,
      "learning_rate": 1.4368000000000002e-05,
      "loss": 0.2766,
      "step": 2210
    },
    {
      "epoch": 3.552,
      "grad_norm": 70.87342834472656,
      "learning_rate": 1.4332444444444446e-05,
      "loss": 0.2622,
      "step": 2220
    },
    {
      "epoch": 3.568,
      "grad_norm": 80.17041778564453,
      "learning_rate": 1.429688888888889e-05,
      "loss": 0.2669,
      "step": 2230
    },
    {
      "epoch": 3.584,
      "grad_norm": 51.33781433105469,
      "learning_rate": 1.4261333333333334e-05,
      "loss": 0.1214,
      "step": 2240
    },
    {
      "epoch": 3.6,
      "grad_norm": 8.450504302978516,
      "learning_rate": 1.4225777777777779e-05,
      "loss": 0.1397,
      "step": 2250
    },
    {
      "epoch": 3.616,
      "grad_norm": 71.05221557617188,
      "learning_rate": 1.4190222222222223e-05,
      "loss": 0.3126,
      "step": 2260
    },
    {
      "epoch": 3.632,
      "grad_norm": 107.07088470458984,
      "learning_rate": 1.4154666666666668e-05,
      "loss": 0.2445,
      "step": 2270
    },
    {
      "epoch": 3.648,
      "grad_norm": 129.06072998046875,
      "learning_rate": 1.4119111111111112e-05,
      "loss": 0.343,
      "step": 2280
    },
    {
      "epoch": 3.664,
      "grad_norm": 114.03457641601562,
      "learning_rate": 1.4083555555555556e-05,
      "loss": 0.4042,
      "step": 2290
    },
    {
      "epoch": 3.68,
      "grad_norm": 98.84168243408203,
      "learning_rate": 1.4048000000000002e-05,
      "loss": 0.4231,
      "step": 2300
    },
    {
      "epoch": 3.6959999999999997,
      "grad_norm": 87.2681655883789,
      "learning_rate": 1.4012444444444446e-05,
      "loss": 0.3805,
      "step": 2310
    },
    {
      "epoch": 3.7119999999999997,
      "grad_norm": 8.487804412841797,
      "learning_rate": 1.3976888888888891e-05,
      "loss": 0.1977,
      "step": 2320
    },
    {
      "epoch": 3.7279999999999998,
      "grad_norm": 57.447540283203125,
      "learning_rate": 1.3941333333333333e-05,
      "loss": 0.3665,
      "step": 2330
    },
    {
      "epoch": 3.7439999999999998,
      "grad_norm": 21.477903366088867,
      "learning_rate": 1.3905777777777779e-05,
      "loss": 0.1807,
      "step": 2340
    },
    {
      "epoch": 3.76,
      "grad_norm": 3.5772314071655273,
      "learning_rate": 1.3870222222222223e-05,
      "loss": 0.2618,
      "step": 2350
    },
    {
      "epoch": 3.776,
      "grad_norm": 71.77428436279297,
      "learning_rate": 1.3834666666666668e-05,
      "loss": 0.1927,
      "step": 2360
    },
    {
      "epoch": 3.792,
      "grad_norm": 115.86641693115234,
      "learning_rate": 1.3799111111111112e-05,
      "loss": 0.2278,
      "step": 2370
    },
    {
      "epoch": 3.808,
      "grad_norm": 32.912540435791016,
      "learning_rate": 1.3763555555555556e-05,
      "loss": 0.3169,
      "step": 2380
    },
    {
      "epoch": 3.824,
      "grad_norm": 90.283447265625,
      "learning_rate": 1.3728000000000001e-05,
      "loss": 0.4014,
      "step": 2390
    },
    {
      "epoch": 3.84,
      "grad_norm": 48.976383209228516,
      "learning_rate": 1.3692444444444445e-05,
      "loss": 0.3318,
      "step": 2400
    },
    {
      "epoch": 3.856,
      "grad_norm": 8.107906341552734,
      "learning_rate": 1.365688888888889e-05,
      "loss": 0.0932,
      "step": 2410
    },
    {
      "epoch": 3.872,
      "grad_norm": 77.96946716308594,
      "learning_rate": 1.3621333333333335e-05,
      "loss": 0.1904,
      "step": 2420
    },
    {
      "epoch": 3.888,
      "grad_norm": 31.673736572265625,
      "learning_rate": 1.358577777777778e-05,
      "loss": 0.3834,
      "step": 2430
    },
    {
      "epoch": 3.904,
      "grad_norm": 132.92593383789062,
      "learning_rate": 1.3550222222222222e-05,
      "loss": 0.3153,
      "step": 2440
    },
    {
      "epoch": 3.92,
      "grad_norm": 48.49889373779297,
      "learning_rate": 1.3514666666666668e-05,
      "loss": 0.2576,
      "step": 2450
    },
    {
      "epoch": 3.936,
      "grad_norm": 34.92246627807617,
      "learning_rate": 1.3479111111111112e-05,
      "loss": 0.16,
      "step": 2460
    },
    {
      "epoch": 3.952,
      "grad_norm": 21.563865661621094,
      "learning_rate": 1.3443555555555556e-05,
      "loss": 0.1484,
      "step": 2470
    },
    {
      "epoch": 3.968,
      "grad_norm": 78.57711029052734,
      "learning_rate": 1.3408000000000001e-05,
      "loss": 0.2304,
      "step": 2480
    },
    {
      "epoch": 3.984,
      "grad_norm": 18.119483947753906,
      "learning_rate": 1.3372444444444445e-05,
      "loss": 0.2055,
      "step": 2490
    },
    {
      "epoch": 4.0,
      "grad_norm": 121.16084289550781,
      "learning_rate": 1.333688888888889e-05,
      "loss": 0.1915,
      "step": 2500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9566666666666667,
      "eval_f1": 0.9586420284715019,
      "eval_loss": 0.13208819925785065,
      "eval_runtime": 25.6321,
      "eval_samples_per_second": 58.52,
      "eval_steps_per_second": 14.63,
      "step": 2500
    },
    {
      "epoch": 4.016,
      "grad_norm": 67.88136291503906,
      "learning_rate": 1.3301333333333334e-05,
      "loss": 0.1068,
      "step": 2510
    },
    {
      "epoch": 4.032,
      "grad_norm": 74.76669311523438,
      "learning_rate": 1.326577777777778e-05,
      "loss": 0.2202,
      "step": 2520
    },
    {
      "epoch": 4.048,
      "grad_norm": 78.5439453125,
      "learning_rate": 1.3230222222222224e-05,
      "loss": 0.257,
      "step": 2530
    },
    {
      "epoch": 4.064,
      "grad_norm": 59.01443099975586,
      "learning_rate": 1.3194666666666668e-05,
      "loss": 0.1632,
      "step": 2540
    },
    {
      "epoch": 4.08,
      "grad_norm": 42.725563049316406,
      "learning_rate": 1.3159111111111111e-05,
      "loss": 0.1885,
      "step": 2550
    },
    {
      "epoch": 4.096,
      "grad_norm": 60.14830017089844,
      "learning_rate": 1.3123555555555557e-05,
      "loss": 0.2077,
      "step": 2560
    },
    {
      "epoch": 4.112,
      "grad_norm": 74.96038055419922,
      "learning_rate": 1.3088e-05,
      "loss": 0.261,
      "step": 2570
    },
    {
      "epoch": 4.128,
      "grad_norm": 97.1154556274414,
      "learning_rate": 1.3052444444444445e-05,
      "loss": 0.3009,
      "step": 2580
    },
    {
      "epoch": 4.144,
      "grad_norm": 11.91926097869873,
      "learning_rate": 1.301688888888889e-05,
      "loss": 0.185,
      "step": 2590
    },
    {
      "epoch": 4.16,
      "grad_norm": 29.406475067138672,
      "learning_rate": 1.2981333333333334e-05,
      "loss": 0.2454,
      "step": 2600
    },
    {
      "epoch": 4.176,
      "grad_norm": 4.042813777923584,
      "learning_rate": 1.294577777777778e-05,
      "loss": 0.1601,
      "step": 2610
    },
    {
      "epoch": 4.192,
      "grad_norm": 99.75011444091797,
      "learning_rate": 1.2910222222222223e-05,
      "loss": 0.2583,
      "step": 2620
    },
    {
      "epoch": 4.208,
      "grad_norm": 72.59245300292969,
      "learning_rate": 1.2874666666666669e-05,
      "loss": 0.3869,
      "step": 2630
    },
    {
      "epoch": 4.224,
      "grad_norm": 79.0741195678711,
      "learning_rate": 1.2839111111111111e-05,
      "loss": 0.2198,
      "step": 2640
    },
    {
      "epoch": 4.24,
      "grad_norm": 31.89919662475586,
      "learning_rate": 1.2803555555555557e-05,
      "loss": 0.1197,
      "step": 2650
    },
    {
      "epoch": 4.256,
      "grad_norm": 94.28645324707031,
      "learning_rate": 1.2768e-05,
      "loss": 0.2221,
      "step": 2660
    },
    {
      "epoch": 4.272,
      "grad_norm": 47.83198547363281,
      "learning_rate": 1.2732444444444444e-05,
      "loss": 0.2703,
      "step": 2670
    },
    {
      "epoch": 4.288,
      "grad_norm": 26.954627990722656,
      "learning_rate": 1.269688888888889e-05,
      "loss": 0.1873,
      "step": 2680
    },
    {
      "epoch": 4.304,
      "grad_norm": 19.696392059326172,
      "learning_rate": 1.2661333333333334e-05,
      "loss": 0.1055,
      "step": 2690
    },
    {
      "epoch": 4.32,
      "grad_norm": 85.9021224975586,
      "learning_rate": 1.262577777777778e-05,
      "loss": 0.2359,
      "step": 2700
    },
    {
      "epoch": 4.336,
      "grad_norm": 69.51016235351562,
      "learning_rate": 1.2590222222222223e-05,
      "loss": 0.3632,
      "step": 2710
    },
    {
      "epoch": 4.352,
      "grad_norm": 44.004737854003906,
      "learning_rate": 1.2554666666666669e-05,
      "loss": 0.2787,
      "step": 2720
    },
    {
      "epoch": 4.368,
      "grad_norm": 14.063050270080566,
      "learning_rate": 1.2519111111111112e-05,
      "loss": 0.2244,
      "step": 2730
    },
    {
      "epoch": 4.384,
      "grad_norm": 34.73302459716797,
      "learning_rate": 1.2483555555555558e-05,
      "loss": 0.2162,
      "step": 2740
    },
    {
      "epoch": 4.4,
      "grad_norm": 3.56476092338562,
      "learning_rate": 1.2448e-05,
      "loss": 0.1797,
      "step": 2750
    },
    {
      "epoch": 4.416,
      "grad_norm": 57.786460876464844,
      "learning_rate": 1.2412444444444444e-05,
      "loss": 0.2107,
      "step": 2760
    },
    {
      "epoch": 4.432,
      "grad_norm": 70.37076568603516,
      "learning_rate": 1.237688888888889e-05,
      "loss": 0.2161,
      "step": 2770
    },
    {
      "epoch": 4.448,
      "grad_norm": 29.15947151184082,
      "learning_rate": 1.2341333333333333e-05,
      "loss": 0.1679,
      "step": 2780
    },
    {
      "epoch": 4.464,
      "grad_norm": 46.475250244140625,
      "learning_rate": 1.2305777777777779e-05,
      "loss": 0.1782,
      "step": 2790
    },
    {
      "epoch": 4.48,
      "grad_norm": 31.443790435791016,
      "learning_rate": 1.2270222222222223e-05,
      "loss": 0.1657,
      "step": 2800
    },
    {
      "epoch": 4.496,
      "grad_norm": 51.556819915771484,
      "learning_rate": 1.2234666666666668e-05,
      "loss": 0.3409,
      "step": 2810
    },
    {
      "epoch": 4.5120000000000005,
      "grad_norm": 3.3252880573272705,
      "learning_rate": 1.2199111111111112e-05,
      "loss": 0.1991,
      "step": 2820
    },
    {
      "epoch": 4.5280000000000005,
      "grad_norm": 5.360611438751221,
      "learning_rate": 1.2163555555555558e-05,
      "loss": 0.2764,
      "step": 2830
    },
    {
      "epoch": 4.5440000000000005,
      "grad_norm": 7.163545608520508,
      "learning_rate": 1.2128000000000001e-05,
      "loss": 0.1315,
      "step": 2840
    },
    {
      "epoch": 4.5600000000000005,
      "grad_norm": 52.50614547729492,
      "learning_rate": 1.2092444444444444e-05,
      "loss": 0.2633,
      "step": 2850
    },
    {
      "epoch": 4.576,
      "grad_norm": 6.198153972625732,
      "learning_rate": 1.2056888888888889e-05,
      "loss": 0.2934,
      "step": 2860
    },
    {
      "epoch": 4.592,
      "grad_norm": 1.0431843996047974,
      "learning_rate": 1.2021333333333333e-05,
      "loss": 0.1791,
      "step": 2870
    },
    {
      "epoch": 4.608,
      "grad_norm": 85.78390502929688,
      "learning_rate": 1.1985777777777779e-05,
      "loss": 0.3817,
      "step": 2880
    },
    {
      "epoch": 4.624,
      "grad_norm": 59.080360412597656,
      "learning_rate": 1.1950222222222222e-05,
      "loss": 0.4065,
      "step": 2890
    },
    {
      "epoch": 4.64,
      "grad_norm": 20.528331756591797,
      "learning_rate": 1.1914666666666668e-05,
      "loss": 0.1737,
      "step": 2900
    },
    {
      "epoch": 4.656,
      "grad_norm": 41.44429016113281,
      "learning_rate": 1.1879111111111112e-05,
      "loss": 0.2387,
      "step": 2910
    },
    {
      "epoch": 4.672,
      "grad_norm": 0.18879224359989166,
      "learning_rate": 1.1843555555555557e-05,
      "loss": 0.2133,
      "step": 2920
    },
    {
      "epoch": 4.688,
      "grad_norm": 45.2426872253418,
      "learning_rate": 1.1808000000000001e-05,
      "loss": 0.285,
      "step": 2930
    },
    {
      "epoch": 4.704,
      "grad_norm": 61.843971252441406,
      "learning_rate": 1.1772444444444447e-05,
      "loss": 0.2561,
      "step": 2940
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.8159428238868713,
      "learning_rate": 1.1736888888888889e-05,
      "loss": 0.185,
      "step": 2950
    },
    {
      "epoch": 4.736,
      "grad_norm": 70.10686492919922,
      "learning_rate": 1.1701333333333333e-05,
      "loss": 0.1155,
      "step": 2960
    },
    {
      "epoch": 4.752,
      "grad_norm": 22.079153060913086,
      "learning_rate": 1.1665777777777778e-05,
      "loss": 0.197,
      "step": 2970
    },
    {
      "epoch": 4.768,
      "grad_norm": 38.955101013183594,
      "learning_rate": 1.1630222222222222e-05,
      "loss": 0.0747,
      "step": 2980
    },
    {
      "epoch": 4.784,
      "grad_norm": 55.808860778808594,
      "learning_rate": 1.1594666666666668e-05,
      "loss": 0.2218,
      "step": 2990
    },
    {
      "epoch": 4.8,
      "grad_norm": 223.2913360595703,
      "learning_rate": 1.1559111111111111e-05,
      "loss": 0.3266,
      "step": 3000
    },
    {
      "epoch": 4.816,
      "grad_norm": 46.87549591064453,
      "learning_rate": 1.1523555555555557e-05,
      "loss": 0.2688,
      "step": 3010
    },
    {
      "epoch": 4.832,
      "grad_norm": 73.7055892944336,
      "learning_rate": 1.1488e-05,
      "loss": 0.2473,
      "step": 3020
    },
    {
      "epoch": 4.848,
      "grad_norm": 3.0956780910491943,
      "learning_rate": 1.1452444444444446e-05,
      "loss": 0.2409,
      "step": 3030
    },
    {
      "epoch": 4.864,
      "grad_norm": 6.121743679046631,
      "learning_rate": 1.141688888888889e-05,
      "loss": 0.2398,
      "step": 3040
    },
    {
      "epoch": 4.88,
      "grad_norm": 1.6634913682937622,
      "learning_rate": 1.1381333333333336e-05,
      "loss": 0.1823,
      "step": 3050
    },
    {
      "epoch": 4.896,
      "grad_norm": 3.9630935192108154,
      "learning_rate": 1.1345777777777778e-05,
      "loss": 0.1687,
      "step": 3060
    },
    {
      "epoch": 4.912,
      "grad_norm": 41.131324768066406,
      "learning_rate": 1.1310222222222222e-05,
      "loss": 0.1912,
      "step": 3070
    },
    {
      "epoch": 4.928,
      "grad_norm": 121.9698715209961,
      "learning_rate": 1.1274666666666667e-05,
      "loss": 0.5025,
      "step": 3080
    },
    {
      "epoch": 4.944,
      "grad_norm": 58.44524002075195,
      "learning_rate": 1.1239111111111111e-05,
      "loss": 0.286,
      "step": 3090
    },
    {
      "epoch": 4.96,
      "grad_norm": 11.90481948852539,
      "learning_rate": 1.1203555555555557e-05,
      "loss": 0.2558,
      "step": 3100
    },
    {
      "epoch": 4.976,
      "grad_norm": 18.4560546875,
      "learning_rate": 1.1168e-05,
      "loss": 0.2252,
      "step": 3110
    },
    {
      "epoch": 4.992,
      "grad_norm": 28.277318954467773,
      "learning_rate": 1.1132444444444446e-05,
      "loss": 0.1953,
      "step": 3120
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.978,
      "eval_f1": 0.9784349951065628,
      "eval_loss": 0.07715080678462982,
      "eval_runtime": 25.1098,
      "eval_samples_per_second": 59.738,
      "eval_steps_per_second": 14.934,
      "step": 3125
    },
    {
      "epoch": 5.008,
      "grad_norm": 38.60945510864258,
      "learning_rate": 1.109688888888889e-05,
      "loss": 0.0791,
      "step": 3130
    },
    {
      "epoch": 5.024,
      "grad_norm": 0.11121569573879242,
      "learning_rate": 1.1061333333333335e-05,
      "loss": 0.2513,
      "step": 3140
    },
    {
      "epoch": 5.04,
      "grad_norm": 7.318007946014404,
      "learning_rate": 1.102577777777778e-05,
      "loss": 0.2184,
      "step": 3150
    },
    {
      "epoch": 5.056,
      "grad_norm": 73.5037612915039,
      "learning_rate": 1.0990222222222221e-05,
      "loss": 0.2048,
      "step": 3160
    },
    {
      "epoch": 5.072,
      "grad_norm": 6.801938533782959,
      "learning_rate": 1.0954666666666667e-05,
      "loss": 0.2091,
      "step": 3170
    },
    {
      "epoch": 5.088,
      "grad_norm": 1.4947863817214966,
      "learning_rate": 1.091911111111111e-05,
      "loss": 0.1544,
      "step": 3180
    },
    {
      "epoch": 5.104,
      "grad_norm": 2.0187385082244873,
      "learning_rate": 1.0883555555555556e-05,
      "loss": 0.1092,
      "step": 3190
    },
    {
      "epoch": 5.12,
      "grad_norm": 49.457881927490234,
      "learning_rate": 1.0848e-05,
      "loss": 0.2173,
      "step": 3200
    },
    {
      "epoch": 5.136,
      "grad_norm": 4.548677444458008,
      "learning_rate": 1.0812444444444446e-05,
      "loss": 0.3133,
      "step": 3210
    },
    {
      "epoch": 5.152,
      "grad_norm": 23.157283782958984,
      "learning_rate": 1.077688888888889e-05,
      "loss": 0.1977,
      "step": 3220
    },
    {
      "epoch": 5.168,
      "grad_norm": 92.80079650878906,
      "learning_rate": 1.0741333333333335e-05,
      "loss": 0.197,
      "step": 3230
    },
    {
      "epoch": 5.184,
      "grad_norm": 23.261850357055664,
      "learning_rate": 1.0705777777777779e-05,
      "loss": 0.1754,
      "step": 3240
    },
    {
      "epoch": 5.2,
      "grad_norm": 70.62091827392578,
      "learning_rate": 1.0670222222222224e-05,
      "loss": 0.1377,
      "step": 3250
    },
    {
      "epoch": 5.216,
      "grad_norm": 1.145323395729065,
      "learning_rate": 1.0634666666666667e-05,
      "loss": 0.2681,
      "step": 3260
    },
    {
      "epoch": 5.232,
      "grad_norm": 82.71636199951172,
      "learning_rate": 1.059911111111111e-05,
      "loss": 0.1608,
      "step": 3270
    },
    {
      "epoch": 5.248,
      "grad_norm": 121.01233673095703,
      "learning_rate": 1.0563555555555556e-05,
      "loss": 0.3524,
      "step": 3280
    },
    {
      "epoch": 5.264,
      "grad_norm": 67.27264404296875,
      "learning_rate": 1.0528e-05,
      "loss": 0.1628,
      "step": 3290
    },
    {
      "epoch": 5.28,
      "grad_norm": 107.12232208251953,
      "learning_rate": 1.0492444444444445e-05,
      "loss": 0.2571,
      "step": 3300
    },
    {
      "epoch": 5.296,
      "grad_norm": 51.44169998168945,
      "learning_rate": 1.045688888888889e-05,
      "loss": 0.2379,
      "step": 3310
    },
    {
      "epoch": 5.312,
      "grad_norm": 92.96754455566406,
      "learning_rate": 1.0421333333333335e-05,
      "loss": 0.1719,
      "step": 3320
    },
    {
      "epoch": 5.328,
      "grad_norm": 0.3315001130104065,
      "learning_rate": 1.0385777777777779e-05,
      "loss": 0.1194,
      "step": 3330
    },
    {
      "epoch": 5.344,
      "grad_norm": 0.11354901641607285,
      "learning_rate": 1.0350222222222224e-05,
      "loss": 0.2085,
      "step": 3340
    },
    {
      "epoch": 5.36,
      "grad_norm": 43.333709716796875,
      "learning_rate": 1.0314666666666668e-05,
      "loss": 0.1196,
      "step": 3350
    },
    {
      "epoch": 5.376,
      "grad_norm": 75.45565032958984,
      "learning_rate": 1.0279111111111114e-05,
      "loss": 0.1169,
      "step": 3360
    },
    {
      "epoch": 5.392,
      "grad_norm": 26.565641403198242,
      "learning_rate": 1.0243555555555556e-05,
      "loss": 0.1446,
      "step": 3370
    },
    {
      "epoch": 5.408,
      "grad_norm": 43.27265930175781,
      "learning_rate": 1.0208e-05,
      "loss": 0.2626,
      "step": 3380
    },
    {
      "epoch": 5.424,
      "grad_norm": 120.37715911865234,
      "learning_rate": 1.0172444444444445e-05,
      "loss": 0.1848,
      "step": 3390
    },
    {
      "epoch": 5.44,
      "grad_norm": 36.994632720947266,
      "learning_rate": 1.0136888888888889e-05,
      "loss": 0.1891,
      "step": 3400
    },
    {
      "epoch": 5.456,
      "grad_norm": 48.42155456542969,
      "learning_rate": 1.0101333333333334e-05,
      "loss": 0.2197,
      "step": 3410
    },
    {
      "epoch": 5.4719999999999995,
      "grad_norm": 3.1949923038482666,
      "learning_rate": 1.0065777777777778e-05,
      "loss": 0.2084,
      "step": 3420
    },
    {
      "epoch": 5.4879999999999995,
      "grad_norm": 125.98743438720703,
      "learning_rate": 1.0030222222222224e-05,
      "loss": 0.1868,
      "step": 3430
    },
    {
      "epoch": 5.504,
      "grad_norm": 96.58470916748047,
      "learning_rate": 9.994666666666668e-06,
      "loss": 0.2648,
      "step": 3440
    },
    {
      "epoch": 5.52,
      "grad_norm": 46.52883529663086,
      "learning_rate": 9.959111111111111e-06,
      "loss": 0.2097,
      "step": 3450
    },
    {
      "epoch": 5.536,
      "grad_norm": 50.802574157714844,
      "learning_rate": 9.923555555555557e-06,
      "loss": 0.3501,
      "step": 3460
    },
    {
      "epoch": 5.552,
      "grad_norm": 56.12847900390625,
      "learning_rate": 9.888000000000001e-06,
      "loss": 0.2818,
      "step": 3470
    },
    {
      "epoch": 5.568,
      "grad_norm": 84.0240249633789,
      "learning_rate": 9.852444444444446e-06,
      "loss": 0.2938,
      "step": 3480
    },
    {
      "epoch": 5.584,
      "grad_norm": 18.963293075561523,
      "learning_rate": 9.81688888888889e-06,
      "loss": 0.2539,
      "step": 3490
    },
    {
      "epoch": 5.6,
      "grad_norm": 20.54253578186035,
      "learning_rate": 9.781333333333334e-06,
      "loss": 0.1566,
      "step": 3500
    },
    {
      "epoch": 5.616,
      "grad_norm": 41.77975082397461,
      "learning_rate": 9.745777777777778e-06,
      "loss": 0.1616,
      "step": 3510
    },
    {
      "epoch": 5.632,
      "grad_norm": 101.94864654541016,
      "learning_rate": 9.710222222222223e-06,
      "loss": 0.2014,
      "step": 3520
    },
    {
      "epoch": 5.648,
      "grad_norm": 66.22476196289062,
      "learning_rate": 9.674666666666667e-06,
      "loss": 0.2032,
      "step": 3530
    },
    {
      "epoch": 5.664,
      "grad_norm": 35.40584182739258,
      "learning_rate": 9.639111111111113e-06,
      "loss": 0.2035,
      "step": 3540
    },
    {
      "epoch": 5.68,
      "grad_norm": 18.508302688598633,
      "learning_rate": 9.603555555555557e-06,
      "loss": 0.2169,
      "step": 3550
    },
    {
      "epoch": 5.696,
      "grad_norm": 2.2750229835510254,
      "learning_rate": 9.568e-06,
      "loss": 0.1242,
      "step": 3560
    },
    {
      "epoch": 5.712,
      "grad_norm": 7.904684543609619,
      "learning_rate": 9.532444444444446e-06,
      "loss": 0.1488,
      "step": 3570
    },
    {
      "epoch": 5.728,
      "grad_norm": 66.34574890136719,
      "learning_rate": 9.49688888888889e-06,
      "loss": 0.3348,
      "step": 3580
    },
    {
      "epoch": 5.744,
      "grad_norm": 2.1702802181243896,
      "learning_rate": 9.461333333333334e-06,
      "loss": 0.169,
      "step": 3590
    },
    {
      "epoch": 5.76,
      "grad_norm": 43.21394348144531,
      "learning_rate": 9.425777777777778e-06,
      "loss": 0.1906,
      "step": 3600
    },
    {
      "epoch": 5.776,
      "grad_norm": 96.57420349121094,
      "learning_rate": 9.390222222222223e-06,
      "loss": 0.2049,
      "step": 3610
    },
    {
      "epoch": 5.792,
      "grad_norm": 60.05061721801758,
      "learning_rate": 9.354666666666667e-06,
      "loss": 0.2975,
      "step": 3620
    },
    {
      "epoch": 5.808,
      "grad_norm": 7.084517478942871,
      "learning_rate": 9.319111111111113e-06,
      "loss": 0.1722,
      "step": 3630
    },
    {
      "epoch": 5.824,
      "grad_norm": 4.0784759521484375,
      "learning_rate": 9.283555555555556e-06,
      "loss": 0.1725,
      "step": 3640
    },
    {
      "epoch": 5.84,
      "grad_norm": 60.38849639892578,
      "learning_rate": 9.248e-06,
      "loss": 0.2305,
      "step": 3650
    },
    {
      "epoch": 5.856,
      "grad_norm": 42.00960159301758,
      "learning_rate": 9.212444444444446e-06,
      "loss": 0.2262,
      "step": 3660
    },
    {
      "epoch": 5.872,
      "grad_norm": 6.289929389953613,
      "learning_rate": 9.17688888888889e-06,
      "loss": 0.1101,
      "step": 3670
    },
    {
      "epoch": 5.888,
      "grad_norm": 46.45234298706055,
      "learning_rate": 9.141333333333333e-06,
      "loss": 0.2282,
      "step": 3680
    },
    {
      "epoch": 5.904,
      "grad_norm": 29.963151931762695,
      "learning_rate": 9.105777777777779e-06,
      "loss": 0.2787,
      "step": 3690
    },
    {
      "epoch": 5.92,
      "grad_norm": 37.11601257324219,
      "learning_rate": 9.070222222222223e-06,
      "loss": 0.2421,
      "step": 3700
    },
    {
      "epoch": 5.936,
      "grad_norm": 21.949438095092773,
      "learning_rate": 9.034666666666667e-06,
      "loss": 0.1956,
      "step": 3710
    },
    {
      "epoch": 5.952,
      "grad_norm": 1.3940507173538208,
      "learning_rate": 8.999111111111112e-06,
      "loss": 0.2005,
      "step": 3720
    },
    {
      "epoch": 5.968,
      "grad_norm": 166.17929077148438,
      "learning_rate": 8.963555555555556e-06,
      "loss": 0.1114,
      "step": 3730
    },
    {
      "epoch": 5.984,
      "grad_norm": 68.92640686035156,
      "learning_rate": 8.928000000000002e-06,
      "loss": 0.2453,
      "step": 3740
    },
    {
      "epoch": 6.0,
      "grad_norm": 40.45570755004883,
      "learning_rate": 8.892444444444445e-06,
      "loss": 0.183,
      "step": 3750
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9773333333333334,
      "eval_f1": 0.9781269999063931,
      "eval_loss": 0.08286113291978836,
      "eval_runtime": 25.561,
      "eval_samples_per_second": 58.683,
      "eval_steps_per_second": 14.671,
      "step": 3750
    },
    {
      "epoch": 6.016,
      "grad_norm": 55.92890167236328,
      "learning_rate": 8.85688888888889e-06,
      "loss": 0.1875,
      "step": 3760
    },
    {
      "epoch": 6.032,
      "grad_norm": 40.81064987182617,
      "learning_rate": 8.821333333333333e-06,
      "loss": 0.1211,
      "step": 3770
    },
    {
      "epoch": 6.048,
      "grad_norm": 37.10798645019531,
      "learning_rate": 8.785777777777779e-06,
      "loss": 0.1888,
      "step": 3780
    },
    {
      "epoch": 6.064,
      "grad_norm": 80.93401336669922,
      "learning_rate": 8.750222222222223e-06,
      "loss": 0.1778,
      "step": 3790
    },
    {
      "epoch": 6.08,
      "grad_norm": 3.9104630947113037,
      "learning_rate": 8.714666666666666e-06,
      "loss": 0.0544,
      "step": 3800
    },
    {
      "epoch": 6.096,
      "grad_norm": 75.7132568359375,
      "learning_rate": 8.679111111111112e-06,
      "loss": 0.2315,
      "step": 3810
    },
    {
      "epoch": 6.112,
      "grad_norm": 3.832777738571167,
      "learning_rate": 8.643555555555556e-06,
      "loss": 0.1239,
      "step": 3820
    },
    {
      "epoch": 6.128,
      "grad_norm": 64.49993896484375,
      "learning_rate": 8.608000000000001e-06,
      "loss": 0.2204,
      "step": 3830
    },
    {
      "epoch": 6.144,
      "grad_norm": 6.4475202560424805,
      "learning_rate": 8.572444444444445e-06,
      "loss": 0.1066,
      "step": 3840
    },
    {
      "epoch": 6.16,
      "grad_norm": 31.276575088500977,
      "learning_rate": 8.53688888888889e-06,
      "loss": 0.1227,
      "step": 3850
    },
    {
      "epoch": 6.176,
      "grad_norm": 24.951356887817383,
      "learning_rate": 8.501333333333334e-06,
      "loss": 0.1531,
      "step": 3860
    },
    {
      "epoch": 6.192,
      "grad_norm": 19.674461364746094,
      "learning_rate": 8.465777777777778e-06,
      "loss": 0.1219,
      "step": 3870
    },
    {
      "epoch": 6.208,
      "grad_norm": 50.07936477661133,
      "learning_rate": 8.430222222222222e-06,
      "loss": 0.1664,
      "step": 3880
    },
    {
      "epoch": 6.224,
      "grad_norm": 60.344425201416016,
      "learning_rate": 8.394666666666668e-06,
      "loss": 0.1654,
      "step": 3890
    },
    {
      "epoch": 6.24,
      "grad_norm": 0.577318012714386,
      "learning_rate": 8.359111111111112e-06,
      "loss": 0.1509,
      "step": 3900
    },
    {
      "epoch": 6.256,
      "grad_norm": 119.1912612915039,
      "learning_rate": 8.323555555555555e-06,
      "loss": 0.2452,
      "step": 3910
    },
    {
      "epoch": 6.272,
      "grad_norm": 92.77301025390625,
      "learning_rate": 8.288000000000001e-06,
      "loss": 0.1129,
      "step": 3920
    },
    {
      "epoch": 6.288,
      "grad_norm": 0.2689651548862457,
      "learning_rate": 8.252444444444445e-06,
      "loss": 0.1511,
      "step": 3930
    },
    {
      "epoch": 6.304,
      "grad_norm": 4.910072326660156,
      "learning_rate": 8.21688888888889e-06,
      "loss": 0.1867,
      "step": 3940
    },
    {
      "epoch": 6.32,
      "grad_norm": 147.96607971191406,
      "learning_rate": 8.181333333333334e-06,
      "loss": 0.1756,
      "step": 3950
    },
    {
      "epoch": 6.336,
      "grad_norm": 3.9301342964172363,
      "learning_rate": 8.145777777777778e-06,
      "loss": 0.0956,
      "step": 3960
    },
    {
      "epoch": 6.352,
      "grad_norm": 4.562084197998047,
      "learning_rate": 8.110222222222222e-06,
      "loss": 0.0985,
      "step": 3970
    },
    {
      "epoch": 6.368,
      "grad_norm": 163.62957763671875,
      "learning_rate": 8.074666666666667e-06,
      "loss": 0.2904,
      "step": 3980
    },
    {
      "epoch": 6.384,
      "grad_norm": 168.3592529296875,
      "learning_rate": 8.039111111111111e-06,
      "loss": 0.2561,
      "step": 3990
    },
    {
      "epoch": 6.4,
      "grad_norm": 70.33308410644531,
      "learning_rate": 8.003555555555557e-06,
      "loss": 0.1576,
      "step": 4000
    },
    {
      "epoch": 6.416,
      "grad_norm": 86.92400360107422,
      "learning_rate": 7.968e-06,
      "loss": 0.2956,
      "step": 4010
    },
    {
      "epoch": 6.432,
      "grad_norm": 100.0543212890625,
      "learning_rate": 7.932444444444444e-06,
      "loss": 0.2009,
      "step": 4020
    },
    {
      "epoch": 6.448,
      "grad_norm": 0.26956602931022644,
      "learning_rate": 7.89688888888889e-06,
      "loss": 0.1287,
      "step": 4030
    },
    {
      "epoch": 6.464,
      "grad_norm": 19.65234375,
      "learning_rate": 7.861333333333334e-06,
      "loss": 0.0629,
      "step": 4040
    },
    {
      "epoch": 6.48,
      "grad_norm": 11.0438871383667,
      "learning_rate": 7.82577777777778e-06,
      "loss": 0.0386,
      "step": 4050
    },
    {
      "epoch": 6.496,
      "grad_norm": 26.837541580200195,
      "learning_rate": 7.790222222222222e-06,
      "loss": 0.0877,
      "step": 4060
    },
    {
      "epoch": 6.5120000000000005,
      "grad_norm": 2.23330020904541,
      "learning_rate": 7.754666666666667e-06,
      "loss": 0.1536,
      "step": 4070
    },
    {
      "epoch": 6.5280000000000005,
      "grad_norm": 103.52494049072266,
      "learning_rate": 7.719111111111111e-06,
      "loss": 0.1594,
      "step": 4080
    },
    {
      "epoch": 6.5440000000000005,
      "grad_norm": 90.99219512939453,
      "learning_rate": 7.683555555555556e-06,
      "loss": 0.1525,
      "step": 4090
    },
    {
      "epoch": 6.5600000000000005,
      "grad_norm": 3.9606099128723145,
      "learning_rate": 7.648e-06,
      "loss": 0.1651,
      "step": 4100
    },
    {
      "epoch": 6.576,
      "grad_norm": 0.9935932159423828,
      "learning_rate": 7.612444444444444e-06,
      "loss": 0.1217,
      "step": 4110
    },
    {
      "epoch": 6.592,
      "grad_norm": 0.7840667366981506,
      "learning_rate": 7.576888888888889e-06,
      "loss": 0.0967,
      "step": 4120
    },
    {
      "epoch": 6.608,
      "grad_norm": 0.795747697353363,
      "learning_rate": 7.5413333333333335e-06,
      "loss": 0.132,
      "step": 4130
    },
    {
      "epoch": 6.624,
      "grad_norm": 5.798128604888916,
      "learning_rate": 7.505777777777778e-06,
      "loss": 0.1085,
      "step": 4140
    },
    {
      "epoch": 6.64,
      "grad_norm": 57.652103424072266,
      "learning_rate": 7.470222222222223e-06,
      "loss": 0.1493,
      "step": 4150
    },
    {
      "epoch": 6.656,
      "grad_norm": 0.1584286391735077,
      "learning_rate": 7.434666666666668e-06,
      "loss": 0.1348,
      "step": 4160
    },
    {
      "epoch": 6.672,
      "grad_norm": 35.631591796875,
      "learning_rate": 7.3991111111111114e-06,
      "loss": 0.1659,
      "step": 4170
    },
    {
      "epoch": 6.688,
      "grad_norm": 36.18688201904297,
      "learning_rate": 7.363555555555556e-06,
      "loss": 0.2248,
      "step": 4180
    },
    {
      "epoch": 6.704,
      "grad_norm": 63.91709899902344,
      "learning_rate": 7.328000000000001e-06,
      "loss": 0.1932,
      "step": 4190
    },
    {
      "epoch": 6.72,
      "grad_norm": 3.0715153217315674,
      "learning_rate": 7.2924444444444455e-06,
      "loss": 0.1612,
      "step": 4200
    },
    {
      "epoch": 6.736,
      "grad_norm": 24.774658203125,
      "learning_rate": 7.25688888888889e-06,
      "loss": 0.1223,
      "step": 4210
    },
    {
      "epoch": 6.752,
      "grad_norm": 0.2880302369594574,
      "learning_rate": 7.221333333333333e-06,
      "loss": 0.0766,
      "step": 4220
    },
    {
      "epoch": 6.768,
      "grad_norm": 193.8282470703125,
      "learning_rate": 7.185777777777778e-06,
      "loss": 0.1071,
      "step": 4230
    },
    {
      "epoch": 6.784,
      "grad_norm": 3.09204363822937,
      "learning_rate": 7.150222222222223e-06,
      "loss": 0.2013,
      "step": 4240
    },
    {
      "epoch": 6.8,
      "grad_norm": 3.475525379180908,
      "learning_rate": 7.114666666666667e-06,
      "loss": 0.1071,
      "step": 4250
    },
    {
      "epoch": 6.816,
      "grad_norm": 105.51129150390625,
      "learning_rate": 7.079111111111112e-06,
      "loss": 0.0891,
      "step": 4260
    },
    {
      "epoch": 6.832,
      "grad_norm": 12.143157958984375,
      "learning_rate": 7.043555555555556e-06,
      "loss": 0.1286,
      "step": 4270
    },
    {
      "epoch": 6.848,
      "grad_norm": 0.539508581161499,
      "learning_rate": 7.0080000000000005e-06,
      "loss": 0.11,
      "step": 4280
    },
    {
      "epoch": 6.864,
      "grad_norm": 16.954879760742188,
      "learning_rate": 6.972444444444445e-06,
      "loss": 0.0895,
      "step": 4290
    },
    {
      "epoch": 6.88,
      "grad_norm": 136.84512329101562,
      "learning_rate": 6.93688888888889e-06,
      "loss": 0.2736,
      "step": 4300
    },
    {
      "epoch": 6.896,
      "grad_norm": 0.7156215906143188,
      "learning_rate": 6.9013333333333346e-06,
      "loss": 0.1564,
      "step": 4310
    },
    {
      "epoch": 6.912,
      "grad_norm": 3.3197035789489746,
      "learning_rate": 6.8657777777777776e-06,
      "loss": 0.0335,
      "step": 4320
    },
    {
      "epoch": 6.928,
      "grad_norm": 22.07130241394043,
      "learning_rate": 6.830222222222222e-06,
      "loss": 0.1674,
      "step": 4330
    },
    {
      "epoch": 6.944,
      "grad_norm": 13.202752113342285,
      "learning_rate": 6.794666666666667e-06,
      "loss": 0.0608,
      "step": 4340
    },
    {
      "epoch": 6.96,
      "grad_norm": 49.284000396728516,
      "learning_rate": 6.759111111111112e-06,
      "loss": 0.1577,
      "step": 4350
    },
    {
      "epoch": 6.976,
      "grad_norm": 7.672852993011475,
      "learning_rate": 6.723555555555556e-06,
      "loss": 0.1058,
      "step": 4360
    },
    {
      "epoch": 6.992,
      "grad_norm": 0.3509444296360016,
      "learning_rate": 6.688e-06,
      "loss": 0.1513,
      "step": 4370
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.972,
      "eval_f1": 0.9735106051850267,
      "eval_loss": 0.11335264146327972,
      "eval_runtime": 25.4218,
      "eval_samples_per_second": 59.004,
      "eval_steps_per_second": 14.751,
      "step": 4375
    },
    {
      "epoch": 7.008,
      "grad_norm": 6.444321155548096,
      "learning_rate": 6.652444444444445e-06,
      "loss": 0.1258,
      "step": 4380
    },
    {
      "epoch": 7.024,
      "grad_norm": 9.443979263305664,
      "learning_rate": 6.6168888888888896e-06,
      "loss": 0.0816,
      "step": 4390
    },
    {
      "epoch": 7.04,
      "grad_norm": 73.74608612060547,
      "learning_rate": 6.581333333333334e-06,
      "loss": 0.1251,
      "step": 4400
    },
    {
      "epoch": 7.056,
      "grad_norm": 1.7877308130264282,
      "learning_rate": 6.545777777777779e-06,
      "loss": 0.0425,
      "step": 4410
    },
    {
      "epoch": 7.072,
      "grad_norm": 52.212074279785156,
      "learning_rate": 6.510222222222222e-06,
      "loss": 0.1424,
      "step": 4420
    },
    {
      "epoch": 7.088,
      "grad_norm": 108.42538452148438,
      "learning_rate": 6.474666666666667e-06,
      "loss": 0.1876,
      "step": 4430
    },
    {
      "epoch": 7.104,
      "grad_norm": 0.7036087512969971,
      "learning_rate": 6.439111111111111e-06,
      "loss": 0.0646,
      "step": 4440
    },
    {
      "epoch": 7.12,
      "grad_norm": 30.391590118408203,
      "learning_rate": 6.403555555555556e-06,
      "loss": 0.1532,
      "step": 4450
    },
    {
      "epoch": 7.136,
      "grad_norm": 30.540075302124023,
      "learning_rate": 6.368000000000001e-06,
      "loss": 0.115,
      "step": 4460
    },
    {
      "epoch": 7.152,
      "grad_norm": 0.48061010241508484,
      "learning_rate": 6.332444444444445e-06,
      "loss": 0.0793,
      "step": 4470
    },
    {
      "epoch": 7.168,
      "grad_norm": 28.764617919921875,
      "learning_rate": 6.296888888888889e-06,
      "loss": 0.071,
      "step": 4480
    },
    {
      "epoch": 7.184,
      "grad_norm": 16.58357810974121,
      "learning_rate": 6.261333333333334e-06,
      "loss": 0.0931,
      "step": 4490
    },
    {
      "epoch": 7.2,
      "grad_norm": 88.44710540771484,
      "learning_rate": 6.225777777777779e-06,
      "loss": 0.2221,
      "step": 4500
    },
    {
      "epoch": 7.216,
      "grad_norm": 16.722288131713867,
      "learning_rate": 6.190222222222223e-06,
      "loss": 0.0601,
      "step": 4510
    },
    {
      "epoch": 7.232,
      "grad_norm": 36.703765869140625,
      "learning_rate": 6.154666666666668e-06,
      "loss": 0.0834,
      "step": 4520
    },
    {
      "epoch": 7.248,
      "grad_norm": 6.038788795471191,
      "learning_rate": 6.119111111111111e-06,
      "loss": 0.107,
      "step": 4530
    },
    {
      "epoch": 7.264,
      "grad_norm": 82.84561920166016,
      "learning_rate": 6.083555555555556e-06,
      "loss": 0.0518,
      "step": 4540
    },
    {
      "epoch": 7.28,
      "grad_norm": 4.832958221435547,
      "learning_rate": 6.048e-06,
      "loss": 0.0873,
      "step": 4550
    },
    {
      "epoch": 7.296,
      "grad_norm": 0.04455806314945221,
      "learning_rate": 6.012444444444445e-06,
      "loss": 0.0628,
      "step": 4560
    },
    {
      "epoch": 7.312,
      "grad_norm": 74.5232162475586,
      "learning_rate": 5.97688888888889e-06,
      "loss": 0.2425,
      "step": 4570
    },
    {
      "epoch": 7.328,
      "grad_norm": 77.59324645996094,
      "learning_rate": 5.941333333333334e-06,
      "loss": 0.1329,
      "step": 4580
    },
    {
      "epoch": 7.344,
      "grad_norm": 14.539669036865234,
      "learning_rate": 5.905777777777778e-06,
      "loss": 0.0915,
      "step": 4590
    },
    {
      "epoch": 7.36,
      "grad_norm": 22.89209747314453,
      "learning_rate": 5.870222222222223e-06,
      "loss": 0.1302,
      "step": 4600
    },
    {
      "epoch": 7.376,
      "grad_norm": 20.032310485839844,
      "learning_rate": 5.834666666666668e-06,
      "loss": 0.1079,
      "step": 4610
    },
    {
      "epoch": 7.392,
      "grad_norm": 92.58062744140625,
      "learning_rate": 5.799111111111112e-06,
      "loss": 0.0798,
      "step": 4620
    },
    {
      "epoch": 7.408,
      "grad_norm": 0.06242356449365616,
      "learning_rate": 5.763555555555555e-06,
      "loss": 0.0412,
      "step": 4630
    },
    {
      "epoch": 7.424,
      "grad_norm": 30.884904861450195,
      "learning_rate": 5.728e-06,
      "loss": 0.1122,
      "step": 4640
    },
    {
      "epoch": 7.44,
      "grad_norm": 64.5287857055664,
      "learning_rate": 5.692444444444445e-06,
      "loss": 0.2161,
      "step": 4650
    },
    {
      "epoch": 7.456,
      "grad_norm": 53.6827392578125,
      "learning_rate": 5.6568888888888894e-06,
      "loss": 0.3298,
      "step": 4660
    },
    {
      "epoch": 7.4719999999999995,
      "grad_norm": 109.19721221923828,
      "learning_rate": 5.621333333333334e-06,
      "loss": 0.1787,
      "step": 4670
    },
    {
      "epoch": 7.4879999999999995,
      "grad_norm": 0.5951263904571533,
      "learning_rate": 5.585777777777778e-06,
      "loss": 0.0459,
      "step": 4680
    },
    {
      "epoch": 7.504,
      "grad_norm": 64.03005981445312,
      "learning_rate": 5.550222222222223e-06,
      "loss": 0.1391,
      "step": 4690
    },
    {
      "epoch": 7.52,
      "grad_norm": 8.298081398010254,
      "learning_rate": 5.514666666666667e-06,
      "loss": 0.0991,
      "step": 4700
    },
    {
      "epoch": 7.536,
      "grad_norm": 142.68936157226562,
      "learning_rate": 5.479111111111112e-06,
      "loss": 0.1358,
      "step": 4710
    },
    {
      "epoch": 7.552,
      "grad_norm": 21.265317916870117,
      "learning_rate": 5.443555555555557e-06,
      "loss": 0.1401,
      "step": 4720
    },
    {
      "epoch": 7.568,
      "grad_norm": 30.693742752075195,
      "learning_rate": 5.408e-06,
      "loss": 0.0769,
      "step": 4730
    },
    {
      "epoch": 7.584,
      "grad_norm": 64.67269134521484,
      "learning_rate": 5.372444444444444e-06,
      "loss": 0.1787,
      "step": 4740
    },
    {
      "epoch": 7.6,
      "grad_norm": 0.0010892553254961967,
      "learning_rate": 5.336888888888889e-06,
      "loss": 0.1196,
      "step": 4750
    },
    {
      "epoch": 7.616,
      "grad_norm": 81.94013977050781,
      "learning_rate": 5.301333333333334e-06,
      "loss": 0.0926,
      "step": 4760
    },
    {
      "epoch": 7.632,
      "grad_norm": 12.81042766571045,
      "learning_rate": 5.2657777777777785e-06,
      "loss": 0.0762,
      "step": 4770
    },
    {
      "epoch": 7.648,
      "grad_norm": 20.876262664794922,
      "learning_rate": 5.230222222222223e-06,
      "loss": 0.1249,
      "step": 4780
    },
    {
      "epoch": 7.664,
      "grad_norm": 3.2425315380096436,
      "learning_rate": 5.194666666666667e-06,
      "loss": 0.0505,
      "step": 4790
    },
    {
      "epoch": 7.68,
      "grad_norm": 54.530662536621094,
      "learning_rate": 5.159111111111112e-06,
      "loss": 0.0492,
      "step": 4800
    },
    {
      "epoch": 7.696,
      "grad_norm": 1.59047269821167,
      "learning_rate": 5.123555555555556e-06,
      "loss": 0.0707,
      "step": 4810
    },
    {
      "epoch": 7.712,
      "grad_norm": 0.6524139046669006,
      "learning_rate": 5.088000000000001e-06,
      "loss": 0.0962,
      "step": 4820
    },
    {
      "epoch": 7.728,
      "grad_norm": 34.15525436401367,
      "learning_rate": 5.052444444444446e-06,
      "loss": 0.0608,
      "step": 4830
    },
    {
      "epoch": 7.744,
      "grad_norm": 25.281831741333008,
      "learning_rate": 5.016888888888889e-06,
      "loss": 0.0812,
      "step": 4840
    },
    {
      "epoch": 7.76,
      "grad_norm": 25.739002227783203,
      "learning_rate": 4.9813333333333335e-06,
      "loss": 0.1723,
      "step": 4850
    },
    {
      "epoch": 7.776,
      "grad_norm": 0.08939272165298462,
      "learning_rate": 4.945777777777778e-06,
      "loss": 0.0563,
      "step": 4860
    },
    {
      "epoch": 7.792,
      "grad_norm": 17.250640869140625,
      "learning_rate": 4.910222222222223e-06,
      "loss": 0.1413,
      "step": 4870
    },
    {
      "epoch": 7.808,
      "grad_norm": 1.4842759370803833,
      "learning_rate": 4.874666666666667e-06,
      "loss": 0.1468,
      "step": 4880
    },
    {
      "epoch": 7.824,
      "grad_norm": 45.6221809387207,
      "learning_rate": 4.839111111111111e-06,
      "loss": 0.1271,
      "step": 4890
    },
    {
      "epoch": 7.84,
      "grad_norm": 75.50129699707031,
      "learning_rate": 4.803555555555556e-06,
      "loss": 0.0428,
      "step": 4900
    },
    {
      "epoch": 7.856,
      "grad_norm": 36.81220626831055,
      "learning_rate": 4.768000000000001e-06,
      "loss": 0.1584,
      "step": 4910
    },
    {
      "epoch": 7.872,
      "grad_norm": 2.676856517791748,
      "learning_rate": 4.7324444444444455e-06,
      "loss": 0.1083,
      "step": 4920
    },
    {
      "epoch": 7.888,
      "grad_norm": 0.38402771949768066,
      "learning_rate": 4.696888888888889e-06,
      "loss": 0.0717,
      "step": 4930
    },
    {
      "epoch": 7.904,
      "grad_norm": 64.51920318603516,
      "learning_rate": 4.661333333333334e-06,
      "loss": 0.1976,
      "step": 4940
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.05417017638683319,
      "learning_rate": 4.625777777777778e-06,
      "loss": 0.0584,
      "step": 4950
    },
    {
      "epoch": 7.936,
      "grad_norm": 39.375465393066406,
      "learning_rate": 4.5902222222222225e-06,
      "loss": 0.0555,
      "step": 4960
    },
    {
      "epoch": 7.952,
      "grad_norm": 0.13842245936393738,
      "learning_rate": 4.554666666666667e-06,
      "loss": 0.1941,
      "step": 4970
    },
    {
      "epoch": 7.968,
      "grad_norm": 0.08617054671049118,
      "learning_rate": 4.519111111111111e-06,
      "loss": 0.1379,
      "step": 4980
    },
    {
      "epoch": 7.984,
      "grad_norm": 1.8630857467651367,
      "learning_rate": 4.483555555555556e-06,
      "loss": 0.1613,
      "step": 4990
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.00913298036903143,
      "learning_rate": 4.4480000000000004e-06,
      "loss": 0.1222,
      "step": 5000
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.972,
      "eval_f1": 0.9728645176079822,
      "eval_loss": 0.11636786162853241,
      "eval_runtime": 25.6737,
      "eval_samples_per_second": 58.426,
      "eval_steps_per_second": 14.606,
      "step": 5000
    },
    {
      "epoch": 8.016,
      "grad_norm": 0.013823838904500008,
      "learning_rate": 4.412444444444445e-06,
      "loss": 0.0628,
      "step": 5010
    },
    {
      "epoch": 8.032,
      "grad_norm": 0.8125199675559998,
      "learning_rate": 4.37688888888889e-06,
      "loss": 0.0465,
      "step": 5020
    },
    {
      "epoch": 8.048,
      "grad_norm": 20.5372371673584,
      "learning_rate": 4.341333333333334e-06,
      "loss": 0.1877,
      "step": 5030
    },
    {
      "epoch": 8.064,
      "grad_norm": 104.10809326171875,
      "learning_rate": 4.305777777777778e-06,
      "loss": 0.0914,
      "step": 5040
    },
    {
      "epoch": 8.08,
      "grad_norm": 8.148750305175781,
      "learning_rate": 4.270222222222223e-06,
      "loss": 0.1,
      "step": 5050
    },
    {
      "epoch": 8.096,
      "grad_norm": 20.803829193115234,
      "learning_rate": 4.234666666666667e-06,
      "loss": 0.2355,
      "step": 5060
    },
    {
      "epoch": 8.112,
      "grad_norm": 5.129018306732178,
      "learning_rate": 4.199111111111112e-06,
      "loss": 0.0997,
      "step": 5070
    },
    {
      "epoch": 8.128,
      "grad_norm": 226.55096435546875,
      "learning_rate": 4.1635555555555554e-06,
      "loss": 0.2214,
      "step": 5080
    },
    {
      "epoch": 8.144,
      "grad_norm": 7.318321704864502,
      "learning_rate": 4.128e-06,
      "loss": 0.0553,
      "step": 5090
    },
    {
      "epoch": 8.16,
      "grad_norm": 21.40049171447754,
      "learning_rate": 4.092444444444445e-06,
      "loss": 0.0479,
      "step": 5100
    },
    {
      "epoch": 8.176,
      "grad_norm": 0.28443557024002075,
      "learning_rate": 4.0568888888888895e-06,
      "loss": 0.0686,
      "step": 5110
    },
    {
      "epoch": 8.192,
      "grad_norm": 0.7155297994613647,
      "learning_rate": 4.021333333333333e-06,
      "loss": 0.0494,
      "step": 5120
    },
    {
      "epoch": 8.208,
      "grad_norm": 3.961090087890625,
      "learning_rate": 3.985777777777778e-06,
      "loss": 0.1256,
      "step": 5130
    },
    {
      "epoch": 8.224,
      "grad_norm": 47.36995315551758,
      "learning_rate": 3.950222222222223e-06,
      "loss": 0.1989,
      "step": 5140
    },
    {
      "epoch": 8.24,
      "grad_norm": 2.0493502616882324,
      "learning_rate": 3.914666666666667e-06,
      "loss": 0.0808,
      "step": 5150
    },
    {
      "epoch": 8.256,
      "grad_norm": 83.30001831054688,
      "learning_rate": 3.879111111111111e-06,
      "loss": 0.0907,
      "step": 5160
    },
    {
      "epoch": 8.272,
      "grad_norm": 0.30890294909477234,
      "learning_rate": 3.843555555555556e-06,
      "loss": 0.0017,
      "step": 5170
    },
    {
      "epoch": 8.288,
      "grad_norm": 10.248932838439941,
      "learning_rate": 3.8080000000000006e-06,
      "loss": 0.0589,
      "step": 5180
    },
    {
      "epoch": 8.304,
      "grad_norm": 49.623043060302734,
      "learning_rate": 3.7724444444444445e-06,
      "loss": 0.1135,
      "step": 5190
    },
    {
      "epoch": 8.32,
      "grad_norm": 0.22713905572891235,
      "learning_rate": 3.736888888888889e-06,
      "loss": 0.0997,
      "step": 5200
    },
    {
      "epoch": 8.336,
      "grad_norm": 0.3724857568740845,
      "learning_rate": 3.7013333333333334e-06,
      "loss": 0.0764,
      "step": 5210
    },
    {
      "epoch": 8.352,
      "grad_norm": 0.576058030128479,
      "learning_rate": 3.665777777777778e-06,
      "loss": 0.0866,
      "step": 5220
    },
    {
      "epoch": 8.368,
      "grad_norm": 67.60527038574219,
      "learning_rate": 3.630222222222223e-06,
      "loss": 0.0999,
      "step": 5230
    },
    {
      "epoch": 8.384,
      "grad_norm": 32.816749572753906,
      "learning_rate": 3.5946666666666667e-06,
      "loss": 0.1252,
      "step": 5240
    },
    {
      "epoch": 8.4,
      "grad_norm": 53.60542678833008,
      "learning_rate": 3.5591111111111114e-06,
      "loss": 0.1626,
      "step": 5250
    },
    {
      "epoch": 8.416,
      "grad_norm": 25.89781951904297,
      "learning_rate": 3.5235555555555556e-06,
      "loss": 0.1803,
      "step": 5260
    },
    {
      "epoch": 8.432,
      "grad_norm": 7.168365001678467,
      "learning_rate": 3.4880000000000003e-06,
      "loss": 0.0965,
      "step": 5270
    },
    {
      "epoch": 8.448,
      "grad_norm": 84.50868225097656,
      "learning_rate": 3.452444444444445e-06,
      "loss": 0.0859,
      "step": 5280
    },
    {
      "epoch": 8.464,
      "grad_norm": 33.135231018066406,
      "learning_rate": 3.416888888888889e-06,
      "loss": 0.1881,
      "step": 5290
    },
    {
      "epoch": 8.48,
      "grad_norm": 27.80241584777832,
      "learning_rate": 3.3813333333333335e-06,
      "loss": 0.0535,
      "step": 5300
    },
    {
      "epoch": 8.496,
      "grad_norm": 5.0428361892700195,
      "learning_rate": 3.345777777777778e-06,
      "loss": 0.1666,
      "step": 5310
    },
    {
      "epoch": 8.512,
      "grad_norm": 0.08426607400178909,
      "learning_rate": 3.3102222222222225e-06,
      "loss": 0.1447,
      "step": 5320
    },
    {
      "epoch": 8.528,
      "grad_norm": 13.224571228027344,
      "learning_rate": 3.274666666666667e-06,
      "loss": 0.2522,
      "step": 5330
    },
    {
      "epoch": 8.544,
      "grad_norm": 36.19147491455078,
      "learning_rate": 3.239111111111111e-06,
      "loss": 0.1451,
      "step": 5340
    },
    {
      "epoch": 8.56,
      "grad_norm": 1.1866250038146973,
      "learning_rate": 3.2035555555555557e-06,
      "loss": 0.1019,
      "step": 5350
    },
    {
      "epoch": 8.576,
      "grad_norm": 61.716758728027344,
      "learning_rate": 3.1680000000000004e-06,
      "loss": 0.1035,
      "step": 5360
    },
    {
      "epoch": 8.592,
      "grad_norm": 7.928038120269775,
      "learning_rate": 3.1324444444444447e-06,
      "loss": 0.1035,
      "step": 5370
    },
    {
      "epoch": 8.608,
      "grad_norm": 1.8415522575378418,
      "learning_rate": 3.0968888888888894e-06,
      "loss": 0.0851,
      "step": 5380
    },
    {
      "epoch": 8.624,
      "grad_norm": 6.056102275848389,
      "learning_rate": 3.0613333333333332e-06,
      "loss": 0.1557,
      "step": 5390
    },
    {
      "epoch": 8.64,
      "grad_norm": 113.13326263427734,
      "learning_rate": 3.025777777777778e-06,
      "loss": 0.1385,
      "step": 5400
    },
    {
      "epoch": 8.656,
      "grad_norm": 52.255775451660156,
      "learning_rate": 2.9902222222222226e-06,
      "loss": 0.0754,
      "step": 5410
    },
    {
      "epoch": 8.672,
      "grad_norm": 0.006745174061506987,
      "learning_rate": 2.954666666666667e-06,
      "loss": 0.0764,
      "step": 5420
    },
    {
      "epoch": 8.688,
      "grad_norm": 2.841777801513672,
      "learning_rate": 2.9191111111111116e-06,
      "loss": 0.1011,
      "step": 5430
    },
    {
      "epoch": 8.704,
      "grad_norm": 2.33146071434021,
      "learning_rate": 2.8835555555555554e-06,
      "loss": 0.1613,
      "step": 5440
    },
    {
      "epoch": 8.72,
      "grad_norm": 36.63135528564453,
      "learning_rate": 2.848e-06,
      "loss": 0.0473,
      "step": 5450
    },
    {
      "epoch": 8.736,
      "grad_norm": 9.347567558288574,
      "learning_rate": 2.8124444444444448e-06,
      "loss": 0.0489,
      "step": 5460
    },
    {
      "epoch": 8.752,
      "grad_norm": 35.201026916503906,
      "learning_rate": 2.776888888888889e-06,
      "loss": 0.0275,
      "step": 5470
    },
    {
      "epoch": 8.768,
      "grad_norm": 0.5583624243736267,
      "learning_rate": 2.7413333333333337e-06,
      "loss": 0.0659,
      "step": 5480
    },
    {
      "epoch": 8.784,
      "grad_norm": 33.14691162109375,
      "learning_rate": 2.7057777777777776e-06,
      "loss": 0.0206,
      "step": 5490
    },
    {
      "epoch": 8.8,
      "grad_norm": 58.47487258911133,
      "learning_rate": 2.6702222222222223e-06,
      "loss": 0.0899,
      "step": 5500
    },
    {
      "epoch": 8.816,
      "grad_norm": 0.43246692419052124,
      "learning_rate": 2.634666666666667e-06,
      "loss": 0.1579,
      "step": 5510
    },
    {
      "epoch": 8.832,
      "grad_norm": 0.14556622505187988,
      "learning_rate": 2.5991111111111112e-06,
      "loss": 0.1158,
      "step": 5520
    },
    {
      "epoch": 8.848,
      "grad_norm": 81.55155944824219,
      "learning_rate": 2.563555555555556e-06,
      "loss": 0.1548,
      "step": 5530
    },
    {
      "epoch": 8.864,
      "grad_norm": 0.013808293268084526,
      "learning_rate": 2.5280000000000006e-06,
      "loss": 0.1388,
      "step": 5540
    },
    {
      "epoch": 8.88,
      "grad_norm": 1.4838815927505493,
      "learning_rate": 2.4924444444444445e-06,
      "loss": 0.0083,
      "step": 5550
    },
    {
      "epoch": 8.896,
      "grad_norm": 98.49444580078125,
      "learning_rate": 2.456888888888889e-06,
      "loss": 0.0779,
      "step": 5560
    },
    {
      "epoch": 8.912,
      "grad_norm": 6.951257705688477,
      "learning_rate": 2.4213333333333334e-06,
      "loss": 0.0272,
      "step": 5570
    },
    {
      "epoch": 8.928,
      "grad_norm": 0.1495244801044464,
      "learning_rate": 2.385777777777778e-06,
      "loss": 0.0335,
      "step": 5580
    },
    {
      "epoch": 8.943999999999999,
      "grad_norm": 0.030582094565033913,
      "learning_rate": 2.3502222222222224e-06,
      "loss": 0.072,
      "step": 5590
    },
    {
      "epoch": 8.96,
      "grad_norm": 0.10239086300134659,
      "learning_rate": 2.3146666666666666e-06,
      "loss": 0.0554,
      "step": 5600
    },
    {
      "epoch": 8.975999999999999,
      "grad_norm": 6.669166564941406,
      "learning_rate": 2.2791111111111113e-06,
      "loss": 0.0266,
      "step": 5610
    },
    {
      "epoch": 8.992,
      "grad_norm": 115.25152587890625,
      "learning_rate": 2.2435555555555556e-06,
      "loss": 0.2001,
      "step": 5620
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9773333333333334,
      "eval_f1": 0.9782696872674377,
      "eval_loss": 0.08652861416339874,
      "eval_runtime": 25.2033,
      "eval_samples_per_second": 59.516,
      "eval_steps_per_second": 14.879,
      "step": 5625
    },
    {
      "epoch": 9.008,
      "grad_norm": 6.198025703430176,
      "learning_rate": 2.2080000000000003e-06,
      "loss": 0.0688,
      "step": 5630
    },
    {
      "epoch": 9.024,
      "grad_norm": 0.08385764062404633,
      "learning_rate": 2.1724444444444446e-06,
      "loss": 0.0546,
      "step": 5640
    },
    {
      "epoch": 9.04,
      "grad_norm": 37.93516540527344,
      "learning_rate": 2.1368888888888892e-06,
      "loss": 0.0769,
      "step": 5650
    },
    {
      "epoch": 9.056,
      "grad_norm": 0.6508764028549194,
      "learning_rate": 2.1013333333333335e-06,
      "loss": 0.063,
      "step": 5660
    },
    {
      "epoch": 9.072,
      "grad_norm": 0.00627525057643652,
      "learning_rate": 2.0657777777777778e-06,
      "loss": 0.0713,
      "step": 5670
    },
    {
      "epoch": 9.088,
      "grad_norm": 6.918279647827148,
      "learning_rate": 2.0302222222222225e-06,
      "loss": 0.1266,
      "step": 5680
    },
    {
      "epoch": 9.104,
      "grad_norm": 5.829410552978516,
      "learning_rate": 1.9946666666666667e-06,
      "loss": 0.0075,
      "step": 5690
    },
    {
      "epoch": 9.12,
      "grad_norm": 0.38724929094314575,
      "learning_rate": 1.9591111111111114e-06,
      "loss": 0.0384,
      "step": 5700
    },
    {
      "epoch": 9.136,
      "grad_norm": 21.070451736450195,
      "learning_rate": 1.9235555555555557e-06,
      "loss": 0.0747,
      "step": 5710
    },
    {
      "epoch": 9.152,
      "grad_norm": 36.819156646728516,
      "learning_rate": 1.8880000000000002e-06,
      "loss": 0.0138,
      "step": 5720
    },
    {
      "epoch": 9.168,
      "grad_norm": 0.718610405921936,
      "learning_rate": 1.8524444444444444e-06,
      "loss": 0.035,
      "step": 5730
    },
    {
      "epoch": 9.184,
      "grad_norm": 124.62521362304688,
      "learning_rate": 1.8168888888888891e-06,
      "loss": 0.0578,
      "step": 5740
    },
    {
      "epoch": 9.2,
      "grad_norm": 24.688091278076172,
      "learning_rate": 1.7813333333333336e-06,
      "loss": 0.1136,
      "step": 5750
    },
    {
      "epoch": 9.216,
      "grad_norm": 2.244947910308838,
      "learning_rate": 1.7457777777777779e-06,
      "loss": 0.2043,
      "step": 5760
    },
    {
      "epoch": 9.232,
      "grad_norm": 41.393150329589844,
      "learning_rate": 1.7102222222222224e-06,
      "loss": 0.0379,
      "step": 5770
    },
    {
      "epoch": 9.248,
      "grad_norm": 0.04206838831305504,
      "learning_rate": 1.6746666666666668e-06,
      "loss": 0.0528,
      "step": 5780
    },
    {
      "epoch": 9.264,
      "grad_norm": 0.4184645116329193,
      "learning_rate": 1.6391111111111113e-06,
      "loss": 0.0387,
      "step": 5790
    },
    {
      "epoch": 9.28,
      "grad_norm": 0.005343704950064421,
      "learning_rate": 1.6035555555555558e-06,
      "loss": 0.0581,
      "step": 5800
    },
    {
      "epoch": 9.296,
      "grad_norm": 4.760532379150391,
      "learning_rate": 1.568e-06,
      "loss": 0.0244,
      "step": 5810
    },
    {
      "epoch": 9.312,
      "grad_norm": 168.53097534179688,
      "learning_rate": 1.5324444444444445e-06,
      "loss": 0.0925,
      "step": 5820
    },
    {
      "epoch": 9.328,
      "grad_norm": 0.011192296631634235,
      "learning_rate": 1.496888888888889e-06,
      "loss": 0.0392,
      "step": 5830
    },
    {
      "epoch": 9.344,
      "grad_norm": 76.10388946533203,
      "learning_rate": 1.4613333333333335e-06,
      "loss": 0.0382,
      "step": 5840
    },
    {
      "epoch": 9.36,
      "grad_norm": 1.840432047843933,
      "learning_rate": 1.4257777777777778e-06,
      "loss": 0.0388,
      "step": 5850
    },
    {
      "epoch": 9.376,
      "grad_norm": 12.041976928710938,
      "learning_rate": 1.3902222222222222e-06,
      "loss": 0.0301,
      "step": 5860
    },
    {
      "epoch": 9.392,
      "grad_norm": 0.9841470122337341,
      "learning_rate": 1.354666666666667e-06,
      "loss": 0.1642,
      "step": 5870
    },
    {
      "epoch": 9.408,
      "grad_norm": 123.9576644897461,
      "learning_rate": 1.3191111111111112e-06,
      "loss": 0.0573,
      "step": 5880
    },
    {
      "epoch": 9.424,
      "grad_norm": 0.32323941588401794,
      "learning_rate": 1.2835555555555557e-06,
      "loss": 0.0091,
      "step": 5890
    },
    {
      "epoch": 9.44,
      "grad_norm": 17.422338485717773,
      "learning_rate": 1.248e-06,
      "loss": 0.0385,
      "step": 5900
    },
    {
      "epoch": 9.456,
      "grad_norm": 9.723565101623535,
      "learning_rate": 1.2124444444444446e-06,
      "loss": 0.1448,
      "step": 5910
    },
    {
      "epoch": 9.472,
      "grad_norm": 68.42117309570312,
      "learning_rate": 1.176888888888889e-06,
      "loss": 0.1427,
      "step": 5920
    },
    {
      "epoch": 9.488,
      "grad_norm": 1.3120919466018677,
      "learning_rate": 1.1413333333333334e-06,
      "loss": 0.0665,
      "step": 5930
    },
    {
      "epoch": 9.504,
      "grad_norm": 20.033823013305664,
      "learning_rate": 1.1057777777777779e-06,
      "loss": 0.1087,
      "step": 5940
    },
    {
      "epoch": 9.52,
      "grad_norm": 104.10298156738281,
      "learning_rate": 1.0702222222222223e-06,
      "loss": 0.1248,
      "step": 5950
    },
    {
      "epoch": 9.536,
      "grad_norm": 94.86436462402344,
      "learning_rate": 1.0346666666666668e-06,
      "loss": 0.0534,
      "step": 5960
    },
    {
      "epoch": 9.552,
      "grad_norm": 52.450225830078125,
      "learning_rate": 9.991111111111113e-07,
      "loss": 0.0851,
      "step": 5970
    },
    {
      "epoch": 9.568,
      "grad_norm": 140.80067443847656,
      "learning_rate": 9.635555555555556e-07,
      "loss": 0.0533,
      "step": 5980
    },
    {
      "epoch": 9.584,
      "grad_norm": 0.12079375237226486,
      "learning_rate": 9.28e-07,
      "loss": 0.0779,
      "step": 5990
    },
    {
      "epoch": 9.6,
      "grad_norm": 2.7201499938964844,
      "learning_rate": 8.924444444444445e-07,
      "loss": 0.0224,
      "step": 6000
    },
    {
      "epoch": 9.616,
      "grad_norm": 0.0656299740076065,
      "learning_rate": 8.568888888888889e-07,
      "loss": 0.0276,
      "step": 6010
    },
    {
      "epoch": 9.632,
      "grad_norm": 0.43925511837005615,
      "learning_rate": 8.213333333333334e-07,
      "loss": 0.0124,
      "step": 6020
    },
    {
      "epoch": 9.648,
      "grad_norm": 0.43047311902046204,
      "learning_rate": 7.857777777777778e-07,
      "loss": 0.0309,
      "step": 6030
    },
    {
      "epoch": 9.664,
      "grad_norm": 0.023367440328001976,
      "learning_rate": 7.502222222222223e-07,
      "loss": 0.0081,
      "step": 6040
    },
    {
      "epoch": 9.68,
      "grad_norm": 0.7051363587379456,
      "learning_rate": 7.146666666666667e-07,
      "loss": 0.0099,
      "step": 6050
    },
    {
      "epoch": 9.696,
      "grad_norm": 1.881511926651001,
      "learning_rate": 6.791111111111112e-07,
      "loss": 0.046,
      "step": 6060
    },
    {
      "epoch": 9.712,
      "grad_norm": 13.757627487182617,
      "learning_rate": 6.435555555555556e-07,
      "loss": 0.1202,
      "step": 6070
    },
    {
      "epoch": 9.728,
      "grad_norm": 59.86408996582031,
      "learning_rate": 6.08e-07,
      "loss": 0.1712,
      "step": 6080
    },
    {
      "epoch": 9.744,
      "grad_norm": 18.352689743041992,
      "learning_rate": 5.724444444444445e-07,
      "loss": 0.0606,
      "step": 6090
    },
    {
      "epoch": 9.76,
      "grad_norm": 0.10592364519834518,
      "learning_rate": 5.368888888888889e-07,
      "loss": 0.063,
      "step": 6100
    },
    {
      "epoch": 9.776,
      "grad_norm": 0.009613972157239914,
      "learning_rate": 5.013333333333334e-07,
      "loss": 0.0522,
      "step": 6110
    },
    {
      "epoch": 9.792,
      "grad_norm": 32.441123962402344,
      "learning_rate": 4.6577777777777785e-07,
      "loss": 0.0699,
      "step": 6120
    },
    {
      "epoch": 9.808,
      "grad_norm": 3.635063648223877,
      "learning_rate": 4.302222222222223e-07,
      "loss": 0.0856,
      "step": 6130
    },
    {
      "epoch": 9.824,
      "grad_norm": 59.86997985839844,
      "learning_rate": 3.9466666666666665e-07,
      "loss": 0.0339,
      "step": 6140
    },
    {
      "epoch": 9.84,
      "grad_norm": 85.35664367675781,
      "learning_rate": 3.5911111111111113e-07,
      "loss": 0.0553,
      "step": 6150
    },
    {
      "epoch": 9.856,
      "grad_norm": 5.168771743774414,
      "learning_rate": 3.2355555555555556e-07,
      "loss": 0.0423,
      "step": 6160
    },
    {
      "epoch": 9.872,
      "grad_norm": 72.00662231445312,
      "learning_rate": 2.8800000000000004e-07,
      "loss": 0.0767,
      "step": 6170
    },
    {
      "epoch": 9.888,
      "grad_norm": 1.257222056388855,
      "learning_rate": 2.5244444444444446e-07,
      "loss": 0.021,
      "step": 6180
    },
    {
      "epoch": 9.904,
      "grad_norm": 0.03861124441027641,
      "learning_rate": 2.1688888888888892e-07,
      "loss": 0.101,
      "step": 6190
    },
    {
      "epoch": 9.92,
      "grad_norm": 60.662452697753906,
      "learning_rate": 1.8133333333333337e-07,
      "loss": 0.0256,
      "step": 6200
    },
    {
      "epoch": 9.936,
      "grad_norm": 57.533695220947266,
      "learning_rate": 1.457777777777778e-07,
      "loss": 0.0846,
      "step": 6210
    },
    {
      "epoch": 9.952,
      "grad_norm": 0.019413741305470467,
      "learning_rate": 1.1022222222222222e-07,
      "loss": 0.0633,
      "step": 6220
    },
    {
      "epoch": 9.968,
      "grad_norm": 0.07436740398406982,
      "learning_rate": 7.466666666666667e-08,
      "loss": 0.0487,
      "step": 6230
    },
    {
      "epoch": 9.984,
      "grad_norm": 0.010338145308196545,
      "learning_rate": 3.911111111111111e-08,
      "loss": 0.0035,
      "step": 6240
    },
    {
      "epoch": 10.0,
      "grad_norm": 47.08232498168945,
      "learning_rate": 3.555555555555556e-09,
      "loss": 0.0618,
      "step": 6250
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.98,
      "eval_f1": 0.9809426117559896,
      "eval_loss": 0.09235712885856628,
      "eval_runtime": 25.2279,
      "eval_samples_per_second": 59.458,
      "eval_steps_per_second": 14.864,
      "step": 6250
    },
    {
      "epoch": 10.0,
      "step": 6250,
      "total_flos": 1.02317615087616e+19,
      "train_loss": 0.2805785644757748,
      "train_runtime": 4473.2535,
      "train_samples_per_second": 22.355,
      "train_steps_per_second": 1.397
    }
  ],
  "logging_steps": 10,
  "max_steps": 6250,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.02317615087616e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}