{
  "best_metric": 0.9737991266375546,
  "best_model_checkpoint": "videomae-base-finetuned-good-gesturePhaseV11/checkpoint-1260",
  "epoch": 9.092857142857143,
  "eval_steps": 500,
  "global_step": 1260,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.007936507936507936,
      "grad_norm": 10.67099666595459,
      "learning_rate": 7.936507936507937e-07,
      "loss": 1.6937,
      "step": 10
    },
    {
      "epoch": 0.015873015873015872,
      "grad_norm": 10.839055061340332,
      "learning_rate": 1.5873015873015873e-06,
      "loss": 1.6052,
      "step": 20
    },
    {
      "epoch": 0.023809523809523808,
      "grad_norm": 10.671605110168457,
      "learning_rate": 2.380952380952381e-06,
      "loss": 1.5128,
      "step": 30
    },
    {
      "epoch": 0.031746031746031744,
      "grad_norm": 10.129815101623535,
      "learning_rate": 3.1746031746031746e-06,
      "loss": 1.3375,
      "step": 40
    },
    {
      "epoch": 0.03968253968253968,
      "grad_norm": 7.233930587768555,
      "learning_rate": 3.968253968253968e-06,
      "loss": 1.1729,
      "step": 50
    },
    {
      "epoch": 0.047619047619047616,
      "grad_norm": 7.941068172454834,
      "learning_rate": 4.761904761904762e-06,
      "loss": 1.0557,
      "step": 60
    },
    {
      "epoch": 0.05555555555555555,
      "grad_norm": 5.641613483428955,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.8538,
      "step": 70
    },
    {
      "epoch": 0.06349206349206349,
      "grad_norm": 11.44336986541748,
      "learning_rate": 6.349206349206349e-06,
      "loss": 0.9715,
      "step": 80
    },
    {
      "epoch": 0.07142857142857142,
      "grad_norm": 6.217037200927734,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 0.9786,
      "step": 90
    },
    {
      "epoch": 0.07936507936507936,
      "grad_norm": 8.418871879577637,
      "learning_rate": 7.936507936507936e-06,
      "loss": 0.8612,
      "step": 100
    },
    {
      "epoch": 0.0873015873015873,
      "grad_norm": 8.026611328125,
      "learning_rate": 8.730158730158731e-06,
      "loss": 0.9033,
      "step": 110
    },
    {
      "epoch": 0.09523809523809523,
      "grad_norm": 5.777014255523682,
      "learning_rate": 9.523809523809525e-06,
      "loss": 1.2097,
      "step": 120
    },
    {
      "epoch": 0.1007936507936508,
      "eval_accuracy": 0.6899563318777293,
      "eval_accuracy_hold": 0.0,
      "eval_accuracy_preparation": 1.0,
      "eval_accuracy_recovery": 0.0,
      "eval_accuracy_stroke": 0.0,
      "eval_accuracy_unknown": 0.0,
      "eval_loss": 1.0243256092071533,
      "eval_runtime": 22.3266,
      "eval_samples_per_second": 10.257,
      "eval_steps_per_second": 1.299,
      "step": 127
    },
    {
      "epoch": 1.0023809523809524,
      "grad_norm": 4.667547225952148,
      "learning_rate": 9.9647266313933e-06,
      "loss": 0.9759,
      "step": 130
    },
    {
      "epoch": 1.0103174603174603,
      "grad_norm": 6.570101261138916,
      "learning_rate": 9.876543209876543e-06,
      "loss": 0.8687,
      "step": 140
    },
    {
      "epoch": 1.0182539682539682,
      "grad_norm": 3.454606533050537,
      "learning_rate": 9.788359788359789e-06,
      "loss": 0.9626,
      "step": 150
    },
    {
      "epoch": 1.026190476190476,
      "grad_norm": 5.504855632781982,
      "learning_rate": 9.700176366843034e-06,
      "loss": 0.8062,
      "step": 160
    },
    {
      "epoch": 1.0341269841269842,
      "grad_norm": 11.000463485717773,
      "learning_rate": 9.61199294532628e-06,
      "loss": 0.8868,
      "step": 170
    },
    {
      "epoch": 1.0420634920634921,
      "grad_norm": 9.436511039733887,
      "learning_rate": 9.523809523809525e-06,
      "loss": 0.7972,
      "step": 180
    },
    {
      "epoch": 1.05,
      "grad_norm": 4.840676784515381,
      "learning_rate": 9.43562610229277e-06,
      "loss": 0.9835,
      "step": 190
    },
    {
      "epoch": 1.057936507936508,
      "grad_norm": 5.491481781005859,
      "learning_rate": 9.347442680776014e-06,
      "loss": 0.8156,
      "step": 200
    },
    {
      "epoch": 1.0658730158730159,
      "grad_norm": 9.045355796813965,
      "learning_rate": 9.25925925925926e-06,
      "loss": 0.9068,
      "step": 210
    },
    {
      "epoch": 1.0738095238095238,
      "grad_norm": 6.567234516143799,
      "learning_rate": 9.171075837742504e-06,
      "loss": 0.9518,
      "step": 220
    },
    {
      "epoch": 1.0817460317460317,
      "grad_norm": 4.041014194488525,
      "learning_rate": 9.08289241622575e-06,
      "loss": 0.8425,
      "step": 230
    },
    {
      "epoch": 1.0896825396825396,
      "grad_norm": 5.872761249542236,
      "learning_rate": 8.994708994708995e-06,
      "loss": 0.7499,
      "step": 240
    },
    {
      "epoch": 1.0976190476190477,
      "grad_norm": 5.4285888671875,
      "learning_rate": 8.90652557319224e-06,
      "loss": 0.8,
      "step": 250
    },
    {
      "epoch": 1.1007936507936509,
      "eval_accuracy": 0.6899563318777293,
      "eval_accuracy_hold": 0.0,
      "eval_accuracy_preparation": 1.0,
      "eval_accuracy_recovery": 0.0,
      "eval_accuracy_stroke": 0.0,
      "eval_accuracy_unknown": 0.0,
      "eval_loss": 0.8716774582862854,
      "eval_runtime": 18.6999,
      "eval_samples_per_second": 12.246,
      "eval_steps_per_second": 1.551,
      "step": 254
    },
    {
      "epoch": 2.0047619047619047,
      "grad_norm": 7.916933536529541,
      "learning_rate": 8.818342151675486e-06,
      "loss": 0.8939,
      "step": 260
    },
    {
      "epoch": 2.0126984126984127,
      "grad_norm": 10.032241821289062,
      "learning_rate": 8.730158730158731e-06,
      "loss": 0.9507,
      "step": 270
    },
    {
      "epoch": 2.0206349206349206,
      "grad_norm": 5.500053405761719,
      "learning_rate": 8.641975308641975e-06,
      "loss": 0.7524,
      "step": 280
    },
    {
      "epoch": 2.0285714285714285,
      "grad_norm": 2.9224343299865723,
      "learning_rate": 8.553791887125222e-06,
      "loss": 0.8033,
      "step": 290
    },
    {
      "epoch": 2.0365079365079364,
      "grad_norm": 9.924591064453125,
      "learning_rate": 8.465608465608466e-06,
      "loss": 0.7143,
      "step": 300
    },
    {
      "epoch": 2.0444444444444443,
      "grad_norm": 6.131772518157959,
      "learning_rate": 8.377425044091711e-06,
      "loss": 0.8056,
      "step": 310
    },
    {
      "epoch": 2.052380952380952,
      "grad_norm": 5.074908256530762,
      "learning_rate": 8.289241622574956e-06,
      "loss": 0.7329,
      "step": 320
    },
    {
      "epoch": 2.06031746031746,
      "grad_norm": 6.0842061042785645,
      "learning_rate": 8.201058201058202e-06,
      "loss": 0.5516,
      "step": 330
    },
    {
      "epoch": 2.0682539682539685,
      "grad_norm": 6.810462951660156,
      "learning_rate": 8.112874779541447e-06,
      "loss": 0.6635,
      "step": 340
    },
    {
      "epoch": 2.0761904761904764,
      "grad_norm": 8.190900802612305,
      "learning_rate": 8.024691358024692e-06,
      "loss": 0.707,
      "step": 350
    },
    {
      "epoch": 2.0841269841269843,
      "grad_norm": 6.851507663726807,
      "learning_rate": 7.936507936507936e-06,
      "loss": 0.8107,
      "step": 360
    },
    {
      "epoch": 2.092063492063492,
      "grad_norm": 10.986032485961914,
      "learning_rate": 7.848324514991183e-06,
      "loss": 0.5181,
      "step": 370
    },
    {
      "epoch": 2.1,
      "grad_norm": 7.873786926269531,
      "learning_rate": 7.760141093474427e-06,
      "loss": 0.5199,
      "step": 380
    },
    {
      "epoch": 2.100793650793651,
      "eval_accuracy": 0.7729257641921398,
      "eval_accuracy_hold": 0.0,
      "eval_accuracy_preparation": 0.9810126582278481,
      "eval_accuracy_recovery": 0.17647058823529413,
      "eval_accuracy_stroke": 0.0,
      "eval_accuracy_unknown": 0.8636363636363636,
      "eval_loss": 0.6724938750267029,
      "eval_runtime": 17.1616,
      "eval_samples_per_second": 13.344,
      "eval_steps_per_second": 1.69,
      "step": 381
    },
    {
      "epoch": 3.007142857142857,
      "grad_norm": 9.23725700378418,
      "learning_rate": 7.671957671957672e-06,
      "loss": 0.5796,
      "step": 390
    },
    {
      "epoch": 3.015079365079365,
      "grad_norm": 14.487571716308594,
      "learning_rate": 7.583774250440918e-06,
      "loss": 0.6356,
      "step": 400
    },
    {
      "epoch": 3.023015873015873,
      "grad_norm": 8.111766815185547,
      "learning_rate": 7.495590828924163e-06,
      "loss": 0.609,
      "step": 410
    },
    {
      "epoch": 3.030952380952381,
      "grad_norm": 10.31395149230957,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.4853,
      "step": 420
    },
    {
      "epoch": 3.0388888888888888,
      "grad_norm": 9.713062286376953,
      "learning_rate": 7.319223985890654e-06,
      "loss": 0.4254,
      "step": 430
    },
    {
      "epoch": 3.0468253968253967,
      "grad_norm": 24.236248016357422,
      "learning_rate": 7.231040564373898e-06,
      "loss": 0.6144,
      "step": 440
    },
    {
      "epoch": 3.0547619047619046,
      "grad_norm": 9.259011268615723,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 0.5036,
      "step": 450
    },
    {
      "epoch": 3.0626984126984125,
      "grad_norm": 9.141166687011719,
      "learning_rate": 7.054673721340388e-06,
      "loss": 0.5061,
      "step": 460
    },
    {
      "epoch": 3.070634920634921,
      "grad_norm": 1.9605706930160522,
      "learning_rate": 6.966490299823634e-06,
      "loss": 0.3654,
      "step": 470
    },
    {
      "epoch": 3.0785714285714287,
      "grad_norm": 23.66200828552246,
      "learning_rate": 6.878306878306879e-06,
      "loss": 0.5331,
      "step": 480
    },
    {
      "epoch": 3.0865079365079366,
      "grad_norm": 17.01007843017578,
      "learning_rate": 6.790123456790124e-06,
      "loss": 0.491,
      "step": 490
    },
    {
      "epoch": 3.0944444444444446,
      "grad_norm": 18.177013397216797,
      "learning_rate": 6.701940035273369e-06,
      "loss": 0.3134,
      "step": 500
    },
    {
      "epoch": 3.100793650793651,
      "eval_accuracy": 0.8427947598253275,
      "eval_accuracy_hold": 0.3076923076923077,
      "eval_accuracy_preparation": 1.0,
      "eval_accuracy_recovery": 0.4117647058823529,
      "eval_accuracy_stroke": 0.0,
      "eval_accuracy_unknown": 0.9090909090909091,
      "eval_loss": 0.47145798802375793,
      "eval_runtime": 20.9475,
      "eval_samples_per_second": 10.932,
      "eval_steps_per_second": 1.384,
      "step": 508
    },
    {
      "epoch": 4.001587301587302,
      "grad_norm": 3.7459118366241455,
      "learning_rate": 6.613756613756615e-06,
      "loss": 0.2951,
      "step": 510
    },
    {
      "epoch": 4.0095238095238095,
      "grad_norm": 1.2948265075683594,
      "learning_rate": 6.525573192239859e-06,
      "loss": 0.3015,
      "step": 520
    },
    {
      "epoch": 4.017460317460317,
      "grad_norm": 28.894683837890625,
      "learning_rate": 6.437389770723105e-06,
      "loss": 0.5519,
      "step": 530
    },
    {
      "epoch": 4.025396825396825,
      "grad_norm": 12.930984497070312,
      "learning_rate": 6.349206349206349e-06,
      "loss": 0.3167,
      "step": 540
    },
    {
      "epoch": 4.033333333333333,
      "grad_norm": 14.674068450927734,
      "learning_rate": 6.2610229276895955e-06,
      "loss": 0.3852,
      "step": 550
    },
    {
      "epoch": 4.041269841269841,
      "grad_norm": 5.561566352844238,
      "learning_rate": 6.17283950617284e-06,
      "loss": 0.2827,
      "step": 560
    },
    {
      "epoch": 4.049206349206349,
      "grad_norm": 2.773526906967163,
      "learning_rate": 6.084656084656085e-06,
      "loss": 0.3025,
      "step": 570
    },
    {
      "epoch": 4.057142857142857,
      "grad_norm": 2.659675121307373,
      "learning_rate": 5.99647266313933e-06,
      "loss": 0.2043,
      "step": 580
    },
    {
      "epoch": 4.065079365079365,
      "grad_norm": 3.98606538772583,
      "learning_rate": 5.908289241622576e-06,
      "loss": 0.3236,
      "step": 590
    },
    {
      "epoch": 4.073015873015873,
      "grad_norm": 27.466781616210938,
      "learning_rate": 5.820105820105821e-06,
      "loss": 0.1979,
      "step": 600
    },
    {
      "epoch": 4.080952380952381,
      "grad_norm": 5.185789108276367,
      "learning_rate": 5.731922398589066e-06,
      "loss": 0.2881,
      "step": 610
    },
    {
      "epoch": 4.088888888888889,
      "grad_norm": 11.606480598449707,
      "learning_rate": 5.6437389770723105e-06,
      "loss": 0.1399,
      "step": 620
    },
    {
      "epoch": 4.0968253968253965,
      "grad_norm": 31.712146759033203,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.1561,
      "step": 630
    },
    {
      "epoch": 4.10079365079365,
      "eval_accuracy": 0.8951965065502183,
      "eval_accuracy_hold": 0.7692307692307693,
      "eval_accuracy_preparation": 1.0,
      "eval_accuracy_recovery": 0.7058823529411765,
      "eval_accuracy_stroke": 0.0,
      "eval_accuracy_unknown": 0.6818181818181818,
      "eval_loss": 0.43630751967430115,
      "eval_runtime": 18.0386,
      "eval_samples_per_second": 12.695,
      "eval_steps_per_second": 1.608,
      "step": 635
    },
    {
      "epoch": 5.003968253968254,
      "grad_norm": 0.8358302116394043,
      "learning_rate": 5.467372134038801e-06,
      "loss": 0.4914,
      "step": 640
    },
    {
      "epoch": 5.011904761904762,
      "grad_norm": 26.23114585876465,
      "learning_rate": 5.3791887125220465e-06,
      "loss": 0.1426,
      "step": 650
    },
    {
      "epoch": 5.01984126984127,
      "grad_norm": 21.137874603271484,
      "learning_rate": 5.291005291005291e-06,
      "loss": 0.3177,
      "step": 660
    },
    {
      "epoch": 5.027777777777778,
      "grad_norm": 13.185637474060059,
      "learning_rate": 5.202821869488537e-06,
      "loss": 0.2641,
      "step": 670
    },
    {
      "epoch": 5.035714285714286,
      "grad_norm": 22.404172897338867,
      "learning_rate": 5.114638447971782e-06,
      "loss": 0.122,
      "step": 680
    },
    {
      "epoch": 5.0436507936507935,
      "grad_norm": 1.1821787357330322,
      "learning_rate": 5.026455026455027e-06,
      "loss": 0.214,
      "step": 690
    },
    {
      "epoch": 5.051587301587301,
      "grad_norm": 0.8503882884979248,
      "learning_rate": 4.938271604938272e-06,
      "loss": 0.3925,
      "step": 700
    },
    {
      "epoch": 5.059523809523809,
      "grad_norm": 26.318893432617188,
      "learning_rate": 4.850088183421517e-06,
      "loss": 0.1443,
      "step": 710
    },
    {
      "epoch": 5.067460317460317,
      "grad_norm": 111.941162109375,
      "learning_rate": 4.761904761904762e-06,
      "loss": 0.1638,
      "step": 720
    },
    {
      "epoch": 5.075396825396825,
      "grad_norm": 27.32931137084961,
      "learning_rate": 4.673721340388007e-06,
      "loss": 0.3608,
      "step": 730
    },
    {
      "epoch": 5.083333333333333,
      "grad_norm": 117.25592041015625,
      "learning_rate": 4.585537918871252e-06,
      "loss": 0.0645,
      "step": 740
    },
    {
      "epoch": 5.091269841269841,
      "grad_norm": 12.846317291259766,
      "learning_rate": 4.497354497354498e-06,
      "loss": 0.2111,
      "step": 750
    },
    {
      "epoch": 5.099206349206349,
      "grad_norm": 45.810054779052734,
      "learning_rate": 4.409171075837743e-06,
      "loss": 0.0429,
      "step": 760
    },
    {
      "epoch": 5.10079365079365,
      "eval_accuracy": 0.9432314410480349,
      "eval_accuracy_hold": 0.8846153846153846,
      "eval_accuracy_preparation": 0.9936708860759493,
      "eval_accuracy_recovery": 0.7058823529411765,
      "eval_accuracy_stroke": 0.5,
      "eval_accuracy_unknown": 0.9545454545454546,
      "eval_loss": 0.2210746556520462,
      "eval_runtime": 17.8804,
      "eval_samples_per_second": 12.807,
      "eval_steps_per_second": 1.622,
      "step": 762
    },
    {
      "epoch": 6.006349206349206,
      "grad_norm": 35.88644027709961,
      "learning_rate": 4.3209876543209875e-06,
      "loss": 0.1086,
      "step": 770
    },
    {
      "epoch": 6.014285714285714,
      "grad_norm": 0.47389915585517883,
      "learning_rate": 4.232804232804233e-06,
      "loss": 0.2244,
      "step": 780
    },
    {
      "epoch": 6.022222222222222,
      "grad_norm": 0.22194349765777588,
      "learning_rate": 4.144620811287478e-06,
      "loss": 0.1538,
      "step": 790
    },
    {
      "epoch": 6.03015873015873,
      "grad_norm": 50.78633499145508,
      "learning_rate": 4.0564373897707236e-06,
      "loss": 0.1374,
      "step": 800
    },
    {
      "epoch": 6.038095238095238,
      "grad_norm": 0.39046162366867065,
      "learning_rate": 3.968253968253968e-06,
      "loss": 0.3255,
      "step": 810
    },
    {
      "epoch": 6.046031746031746,
      "grad_norm": 56.010379791259766,
      "learning_rate": 3.8800705467372134e-06,
      "loss": 0.1618,
      "step": 820
    },
    {
      "epoch": 6.053968253968254,
      "grad_norm": 0.16263973712921143,
      "learning_rate": 3.791887125220459e-06,
      "loss": 0.116,
      "step": 830
    },
    {
      "epoch": 6.061904761904762,
      "grad_norm": 0.23908407986164093,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.1012,
      "step": 840
    },
    {
      "epoch": 6.06984126984127,
      "grad_norm": 53.480140686035156,
      "learning_rate": 3.615520282186949e-06,
      "loss": 0.1767,
      "step": 850
    },
    {
      "epoch": 6.0777777777777775,
      "grad_norm": 0.681148886680603,
      "learning_rate": 3.527336860670194e-06,
      "loss": 0.1364,
      "step": 860
    },
    {
      "epoch": 6.085714285714285,
      "grad_norm": 6.9544267654418945,
      "learning_rate": 3.4391534391534394e-06,
      "loss": 0.0884,
      "step": 870
    },
    {
      "epoch": 6.093650793650793,
      "grad_norm": 0.3266388177871704,
      "learning_rate": 3.3509700176366843e-06,
      "loss": 0.2294,
      "step": 880
    },
    {
      "epoch": 6.10079365079365,
      "eval_accuracy": 0.9475982532751092,
      "eval_accuracy_hold": 0.8846153846153846,
      "eval_accuracy_preparation": 1.0,
      "eval_accuracy_recovery": 0.8823529411764706,
      "eval_accuracy_stroke": 0.16666666666666666,
      "eval_accuracy_unknown": 0.9090909090909091,
      "eval_loss": 0.20944689214229584,
      "eval_runtime": 16.986,
      "eval_samples_per_second": 13.482,
      "eval_steps_per_second": 1.707,
      "step": 889
    },
    {
      "epoch": 7.000793650793651,
      "grad_norm": 4.641857147216797,
      "learning_rate": 3.2627865961199297e-06,
      "loss": 0.1946,
      "step": 890
    },
    {
      "epoch": 7.008730158730159,
      "grad_norm": 33.782859802246094,
      "learning_rate": 3.1746031746031746e-06,
      "loss": 0.0758,
      "step": 900
    },
    {
      "epoch": 7.016666666666667,
      "grad_norm": 0.22343800961971283,
      "learning_rate": 3.08641975308642e-06,
      "loss": 0.0885,
      "step": 910
    },
    {
      "epoch": 7.0246031746031745,
      "grad_norm": 4.360742092132568,
      "learning_rate": 2.998236331569665e-06,
      "loss": 0.016,
      "step": 920
    },
    {
      "epoch": 7.032539682539682,
      "grad_norm": 0.2488667219877243,
      "learning_rate": 2.9100529100529103e-06,
      "loss": 0.1495,
      "step": 930
    },
    {
      "epoch": 7.04047619047619,
      "grad_norm": 28.20427703857422,
      "learning_rate": 2.8218694885361552e-06,
      "loss": 0.1959,
      "step": 940
    },
    {
      "epoch": 7.048412698412698,
      "grad_norm": 8.172640800476074,
      "learning_rate": 2.7336860670194006e-06,
      "loss": 0.0934,
      "step": 950
    },
    {
      "epoch": 7.056349206349206,
      "grad_norm": 9.601517677307129,
      "learning_rate": 2.6455026455026455e-06,
      "loss": 0.2069,
      "step": 960
    },
    {
      "epoch": 7.064285714285714,
      "grad_norm": 17.030954360961914,
      "learning_rate": 2.557319223985891e-06,
      "loss": 0.0529,
      "step": 970
    },
    {
      "epoch": 7.072222222222222,
      "grad_norm": 0.2699105143547058,
      "learning_rate": 2.469135802469136e-06,
      "loss": 0.236,
      "step": 980
    },
    {
      "epoch": 7.08015873015873,
      "grad_norm": 0.7129911184310913,
      "learning_rate": 2.380952380952381e-06,
      "loss": 0.015,
      "step": 990
    },
    {
      "epoch": 7.088095238095238,
      "grad_norm": 20.803991317749023,
      "learning_rate": 2.292768959435626e-06,
      "loss": 0.07,
      "step": 1000
    },
    {
      "epoch": 7.096031746031746,
      "grad_norm": 31.60879135131836,
      "learning_rate": 2.2045855379188715e-06,
      "loss": 0.1214,
      "step": 1010
    },
    {
      "epoch": 7.10079365079365,
      "eval_accuracy": 0.9606986899563319,
      "eval_accuracy_hold": 0.8846153846153846,
      "eval_accuracy_preparation": 0.9936708860759493,
      "eval_accuracy_recovery": 0.8823529411764706,
      "eval_accuracy_stroke": 0.6666666666666666,
      "eval_accuracy_unknown": 0.9545454545454546,
      "eval_loss": 0.15864819288253784,
      "eval_runtime": 198.6859,
      "eval_samples_per_second": 1.153,
      "eval_steps_per_second": 0.146,
      "step": 1016
    },
    {
      "epoch": 8.003174603174603,
      "grad_norm": 0.32036104798316956,
      "learning_rate": 2.1164021164021164e-06,
      "loss": 0.0981,
      "step": 1020
    },
    {
      "epoch": 8.011111111111111,
      "grad_norm": 0.1587836891412735,
      "learning_rate": 2.0282186948853618e-06,
      "loss": 0.1222,
      "step": 1030
    },
    {
      "epoch": 8.019047619047619,
      "grad_norm": 9.007894515991211,
      "learning_rate": 1.9400352733686067e-06,
      "loss": 0.2571,
      "step": 1040
    },
    {
      "epoch": 8.026984126984127,
      "grad_norm": 0.25755736231803894,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 0.0791,
      "step": 1050
    },
    {
      "epoch": 8.034920634920635,
      "grad_norm": 6.88518762588501,
      "learning_rate": 1.763668430335097e-06,
      "loss": 0.0597,
      "step": 1060
    },
    {
      "epoch": 8.042857142857143,
      "grad_norm": 28.681108474731445,
      "learning_rate": 1.6754850088183422e-06,
      "loss": 0.1355,
      "step": 1070
    },
    {
      "epoch": 8.05079365079365,
      "grad_norm": 0.7820481061935425,
      "learning_rate": 1.5873015873015873e-06,
      "loss": 0.0102,
      "step": 1080
    },
    {
      "epoch": 8.058730158730159,
      "grad_norm": 0.11372612416744232,
      "learning_rate": 1.4991181657848325e-06,
      "loss": 0.0969,
      "step": 1090
    },
    {
      "epoch": 8.066666666666666,
      "grad_norm": 0.962800145149231,
      "learning_rate": 1.4109347442680776e-06,
      "loss": 0.0633,
      "step": 1100
    },
    {
      "epoch": 8.074603174603174,
      "grad_norm": 0.29850736260414124,
      "learning_rate": 1.3227513227513228e-06,
      "loss": 0.0743,
      "step": 1110
    },
    {
      "epoch": 8.082539682539682,
      "grad_norm": 0.2638128101825714,
      "learning_rate": 1.234567901234568e-06,
      "loss": 0.0603,
      "step": 1120
    },
    {
      "epoch": 8.09047619047619,
      "grad_norm": 4.3052215576171875,
      "learning_rate": 1.146384479717813e-06,
      "loss": 0.1776,
      "step": 1130
    },
    {
      "epoch": 8.098412698412698,
      "grad_norm": 7.065666675567627,
      "learning_rate": 1.0582010582010582e-06,
      "loss": 0.1478,
      "step": 1140
    },
    {
      "epoch": 8.100793650793651,
      "eval_accuracy": 0.9432314410480349,
      "eval_accuracy_hold": 0.8846153846153846,
      "eval_accuracy_preparation": 0.9683544303797469,
      "eval_accuracy_recovery": 0.8823529411764706,
      "eval_accuracy_stroke": 0.6666666666666666,
      "eval_accuracy_unknown": 0.9545454545454546,
      "eval_loss": 0.16072985529899597,
      "eval_runtime": 38.5322,
      "eval_samples_per_second": 5.943,
      "eval_steps_per_second": 0.753,
      "step": 1143
    },
    {
      "epoch": 9.005555555555556,
      "grad_norm": 1.259048342704773,
      "learning_rate": 9.700176366843034e-07,
      "loss": 0.0626,
      "step": 1150
    },
    {
      "epoch": 9.013492063492064,
      "grad_norm": 13.264341354370117,
      "learning_rate": 8.818342151675485e-07,
      "loss": 0.1718,
      "step": 1160
    },
    {
      "epoch": 9.021428571428572,
      "grad_norm": 0.916714608669281,
      "learning_rate": 7.936507936507937e-07,
      "loss": 0.0135,
      "step": 1170
    },
    {
      "epoch": 9.02936507936508,
      "grad_norm": 0.167047917842865,
      "learning_rate": 7.054673721340388e-07,
      "loss": 0.087,
      "step": 1180
    },
    {
      "epoch": 9.037301587301588,
      "grad_norm": 0.09918953478336334,
      "learning_rate": 6.17283950617284e-07,
      "loss": 0.0936,
      "step": 1190
    },
    {
      "epoch": 9.045238095238096,
      "grad_norm": 9.908056259155273,
      "learning_rate": 5.291005291005291e-07,
      "loss": 0.1706,
      "step": 1200
    },
    {
      "epoch": 9.053174603174604,
      "grad_norm": 0.15841275453567505,
      "learning_rate": 4.4091710758377425e-07,
      "loss": 0.0411,
      "step": 1210
    },
    {
      "epoch": 9.061111111111112,
      "grad_norm": 9.066829681396484,
      "learning_rate": 3.527336860670194e-07,
      "loss": 0.0798,
      "step": 1220
    },
    {
      "epoch": 9.06904761904762,
      "grad_norm": 15.856799125671387,
      "learning_rate": 2.6455026455026455e-07,
      "loss": 0.0698,
      "step": 1230
    },
    {
      "epoch": 9.076984126984128,
      "grad_norm": 57.29991149902344,
      "learning_rate": 1.763668430335097e-07,
      "loss": 0.0994,
      "step": 1240
    },
    {
      "epoch": 9.084920634920636,
      "grad_norm": 0.18446692824363708,
      "learning_rate": 8.818342151675485e-08,
      "loss": 0.1576,
      "step": 1250
    },
    {
      "epoch": 9.092857142857143,
      "grad_norm": 0.1289648562669754,
      "learning_rate": 0.0,
      "loss": 0.1156,
      "step": 1260
    },
    {
      "epoch": 9.092857142857143,
      "eval_accuracy": 0.9737991266375546,
      "eval_accuracy_hold": 1.0,
      "eval_accuracy_preparation": 0.9936708860759493,
      "eval_accuracy_recovery": 0.8823529411764706,
      "eval_accuracy_stroke": 0.6666666666666666,
      "eval_accuracy_unknown": 0.9545454545454546,
      "eval_loss": 0.11774494498968124,
      "eval_runtime": 38.5752,
      "eval_samples_per_second": 5.936,
      "eval_steps_per_second": 0.752,
      "step": 1260
    },
    {
      "epoch": 9.092857142857143,
      "step": 1260,
      "total_flos": 1.2549452224921436e+19,
      "train_loss": 0.42608252665589724,
      "train_runtime": 2594.6501,
      "train_samples_per_second": 3.885,
      "train_steps_per_second": 0.486
    },
    {
      "epoch": 9.092857142857143,
      "eval_accuracy": 0.946058091286307,
      "eval_accuracy_hold": 1.0,
      "eval_accuracy_preparation": 0.9685534591194969,
      "eval_accuracy_recovery": 0.8947368421052632,
      "eval_accuracy_stroke": 0.42857142857142855,
      "eval_accuracy_unknown": 0.9285714285714286,
      "eval_loss": 0.19672314822673798,
      "eval_runtime": 33.9773,
      "eval_samples_per_second": 7.093,
      "eval_steps_per_second": 0.912,
      "step": 1260
    },
    {
      "epoch": 9.092857142857143,
      "eval_accuracy": 0.946058091286307,
      "eval_accuracy_hold": 1.0,
      "eval_accuracy_preparation": 0.9685534591194969,
      "eval_accuracy_recovery": 0.8947368421052632,
      "eval_accuracy_stroke": 0.42857142857142855,
      "eval_accuracy_unknown": 0.9285714285714286,
      "eval_loss": 0.19672317802906036,
      "eval_runtime": 30.5504,
      "eval_samples_per_second": 7.889,
      "eval_steps_per_second": 1.015,
      "step": 1260
    }
  ],
  "logging_steps": 10,
  "max_steps": 1260,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "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": 1.2549452224921436e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}