{
  "best_metric": 0.8703819661865999,
  "best_model_checkpoint": "dinov2-base-finetuned-galaxy/checkpoint-2240",
  "epoch": 19.955456570155903,
  "eval_steps": 500,
  "global_step": 2240,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08908685968819599,
      "grad_norm": 41.429115295410156,
      "learning_rate": 2.2321428571428573e-06,
      "loss": 2.5179,
      "step": 10
    },
    {
      "epoch": 0.17817371937639198,
      "grad_norm": 39.681358337402344,
      "learning_rate": 4.464285714285715e-06,
      "loss": 2.0573,
      "step": 20
    },
    {
      "epoch": 0.267260579064588,
      "grad_norm": 43.90978240966797,
      "learning_rate": 6.696428571428572e-06,
      "loss": 1.5858,
      "step": 30
    },
    {
      "epoch": 0.35634743875278396,
      "grad_norm": 58.3934211730957,
      "learning_rate": 8.92857142857143e-06,
      "loss": 1.315,
      "step": 40
    },
    {
      "epoch": 0.44543429844098,
      "grad_norm": 50.093204498291016,
      "learning_rate": 1.1160714285714287e-05,
      "loss": 1.1625,
      "step": 50
    },
    {
      "epoch": 0.534521158129176,
      "grad_norm": 62.2159423828125,
      "learning_rate": 1.3392857142857144e-05,
      "loss": 1.068,
      "step": 60
    },
    {
      "epoch": 0.623608017817372,
      "grad_norm": 71.58743286132812,
      "learning_rate": 1.5625e-05,
      "loss": 0.9487,
      "step": 70
    },
    {
      "epoch": 0.7126948775055679,
      "grad_norm": 71.40558624267578,
      "learning_rate": 1.785714285714286e-05,
      "loss": 1.0095,
      "step": 80
    },
    {
      "epoch": 0.8017817371937639,
      "grad_norm": 100.2014389038086,
      "learning_rate": 2.0089285714285717e-05,
      "loss": 1.0315,
      "step": 90
    },
    {
      "epoch": 0.89086859688196,
      "grad_norm": 40.66311264038086,
      "learning_rate": 2.2321428571428575e-05,
      "loss": 0.8269,
      "step": 100
    },
    {
      "epoch": 0.9799554565701559,
      "grad_norm": 82.63133239746094,
      "learning_rate": 2.455357142857143e-05,
      "loss": 0.8589,
      "step": 110
    },
    {
      "epoch": 0.9977728285077951,
      "eval_accuracy": 0.7507827175954915,
      "eval_loss": 0.7458012104034424,
      "eval_runtime": 21.2216,
      "eval_samples_per_second": 75.253,
      "eval_steps_per_second": 2.356,
      "step": 112
    },
    {
      "epoch": 1.069042316258352,
      "grad_norm": 38.83303451538086,
      "learning_rate": 2.6785714285714288e-05,
      "loss": 0.8078,
      "step": 120
    },
    {
      "epoch": 1.158129175946548,
      "grad_norm": 68.20454406738281,
      "learning_rate": 2.9017857142857146e-05,
      "loss": 0.8021,
      "step": 130
    },
    {
      "epoch": 1.247216035634744,
      "grad_norm": 46.07291030883789,
      "learning_rate": 3.125e-05,
      "loss": 0.8225,
      "step": 140
    },
    {
      "epoch": 1.3363028953229399,
      "grad_norm": 72.90870666503906,
      "learning_rate": 3.348214285714286e-05,
      "loss": 0.8584,
      "step": 150
    },
    {
      "epoch": 1.4253897550111359,
      "grad_norm": 56.07601547241211,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.894,
      "step": 160
    },
    {
      "epoch": 1.5144766146993318,
      "grad_norm": 39.8024787902832,
      "learning_rate": 3.794642857142857e-05,
      "loss": 0.9406,
      "step": 170
    },
    {
      "epoch": 1.6035634743875278,
      "grad_norm": 59.87102127075195,
      "learning_rate": 4.017857142857143e-05,
      "loss": 0.8083,
      "step": 180
    },
    {
      "epoch": 1.692650334075724,
      "grad_norm": 74.46855926513672,
      "learning_rate": 4.2410714285714285e-05,
      "loss": 0.8802,
      "step": 190
    },
    {
      "epoch": 1.7817371937639197,
      "grad_norm": 30.22767448425293,
      "learning_rate": 4.464285714285715e-05,
      "loss": 0.8672,
      "step": 200
    },
    {
      "epoch": 1.8708240534521159,
      "grad_norm": 32.281150817871094,
      "learning_rate": 4.6875e-05,
      "loss": 0.8663,
      "step": 210
    },
    {
      "epoch": 1.9599109131403119,
      "grad_norm": 37.24480056762695,
      "learning_rate": 4.910714285714286e-05,
      "loss": 0.8465,
      "step": 220
    },
    {
      "epoch": 1.9955456570155903,
      "eval_accuracy": 0.7864746399499061,
      "eval_loss": 0.656895637512207,
      "eval_runtime": 21.1029,
      "eval_samples_per_second": 75.677,
      "eval_steps_per_second": 2.369,
      "step": 224
    },
    {
      "epoch": 2.048997772828508,
      "grad_norm": 40.045684814453125,
      "learning_rate": 4.985119047619048e-05,
      "loss": 0.7752,
      "step": 230
    },
    {
      "epoch": 2.138084632516704,
      "grad_norm": 26.04545783996582,
      "learning_rate": 4.960317460317461e-05,
      "loss": 0.7703,
      "step": 240
    },
    {
      "epoch": 2.2271714922048997,
      "grad_norm": 38.471458435058594,
      "learning_rate": 4.9355158730158735e-05,
      "loss": 0.8756,
      "step": 250
    },
    {
      "epoch": 2.316258351893096,
      "grad_norm": 22.920825958251953,
      "learning_rate": 4.910714285714286e-05,
      "loss": 0.8719,
      "step": 260
    },
    {
      "epoch": 2.4053452115812917,
      "grad_norm": 25.196147918701172,
      "learning_rate": 4.8859126984126984e-05,
      "loss": 0.8446,
      "step": 270
    },
    {
      "epoch": 2.494432071269488,
      "grad_norm": 37.12013626098633,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 0.7929,
      "step": 280
    },
    {
      "epoch": 2.5835189309576836,
      "grad_norm": 40.25604248046875,
      "learning_rate": 4.836309523809524e-05,
      "loss": 0.7952,
      "step": 290
    },
    {
      "epoch": 2.6726057906458798,
      "grad_norm": 33.80072021484375,
      "learning_rate": 4.811507936507937e-05,
      "loss": 0.7557,
      "step": 300
    },
    {
      "epoch": 2.7616926503340755,
      "grad_norm": 18.884248733520508,
      "learning_rate": 4.7867063492063496e-05,
      "loss": 0.8124,
      "step": 310
    },
    {
      "epoch": 2.8507795100222717,
      "grad_norm": 23.225616455078125,
      "learning_rate": 4.761904761904762e-05,
      "loss": 0.71,
      "step": 320
    },
    {
      "epoch": 2.939866369710468,
      "grad_norm": 23.466827392578125,
      "learning_rate": 4.7371031746031745e-05,
      "loss": 0.732,
      "step": 330
    },
    {
      "epoch": 2.9933184855233854,
      "eval_accuracy": 0.7595491546649968,
      "eval_loss": 0.7020854353904724,
      "eval_runtime": 21.064,
      "eval_samples_per_second": 75.816,
      "eval_steps_per_second": 2.374,
      "step": 336
    },
    {
      "epoch": 3.0289532293986636,
      "grad_norm": 23.42888069152832,
      "learning_rate": 4.7123015873015876e-05,
      "loss": 0.7415,
      "step": 340
    },
    {
      "epoch": 3.11804008908686,
      "grad_norm": 35.59011459350586,
      "learning_rate": 4.6875e-05,
      "loss": 0.7201,
      "step": 350
    },
    {
      "epoch": 3.2071269487750556,
      "grad_norm": 39.206329345703125,
      "learning_rate": 4.662698412698413e-05,
      "loss": 0.7227,
      "step": 360
    },
    {
      "epoch": 3.2962138084632517,
      "grad_norm": 25.05638885498047,
      "learning_rate": 4.637896825396826e-05,
      "loss": 0.7214,
      "step": 370
    },
    {
      "epoch": 3.3853006681514475,
      "grad_norm": 19.9650936126709,
      "learning_rate": 4.613095238095239e-05,
      "loss": 0.7479,
      "step": 380
    },
    {
      "epoch": 3.4743875278396437,
      "grad_norm": 30.844806671142578,
      "learning_rate": 4.5882936507936506e-05,
      "loss": 0.7454,
      "step": 390
    },
    {
      "epoch": 3.5634743875278394,
      "grad_norm": 16.321706771850586,
      "learning_rate": 4.563492063492064e-05,
      "loss": 0.7032,
      "step": 400
    },
    {
      "epoch": 3.6525612472160356,
      "grad_norm": 15.168939590454102,
      "learning_rate": 4.538690476190476e-05,
      "loss": 0.6498,
      "step": 410
    },
    {
      "epoch": 3.7416481069042318,
      "grad_norm": 30.41251564025879,
      "learning_rate": 4.5138888888888894e-05,
      "loss": 0.7092,
      "step": 420
    },
    {
      "epoch": 3.8307349665924275,
      "grad_norm": 22.254060745239258,
      "learning_rate": 4.489087301587302e-05,
      "loss": 0.7199,
      "step": 430
    },
    {
      "epoch": 3.9198218262806237,
      "grad_norm": 17.9412841796875,
      "learning_rate": 4.464285714285715e-05,
      "loss": 0.7083,
      "step": 440
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7883531621790858,
      "eval_loss": 0.6093864440917969,
      "eval_runtime": 21.2086,
      "eval_samples_per_second": 75.3,
      "eval_steps_per_second": 2.358,
      "step": 449
    },
    {
      "epoch": 4.008908685968819,
      "grad_norm": 20.122547149658203,
      "learning_rate": 4.439484126984127e-05,
      "loss": 0.6959,
      "step": 450
    },
    {
      "epoch": 4.097995545657016,
      "grad_norm": 20.436920166015625,
      "learning_rate": 4.41468253968254e-05,
      "loss": 0.5657,
      "step": 460
    },
    {
      "epoch": 4.187082405345212,
      "grad_norm": 18.810733795166016,
      "learning_rate": 4.3898809523809523e-05,
      "loss": 0.645,
      "step": 470
    },
    {
      "epoch": 4.276169265033408,
      "grad_norm": 19.18097496032715,
      "learning_rate": 4.3650793650793655e-05,
      "loss": 0.7224,
      "step": 480
    },
    {
      "epoch": 4.365256124721603,
      "grad_norm": 14.759562492370605,
      "learning_rate": 4.340277777777778e-05,
      "loss": 0.6456,
      "step": 490
    },
    {
      "epoch": 4.4543429844097995,
      "grad_norm": 26.570215225219727,
      "learning_rate": 4.315476190476191e-05,
      "loss": 0.7037,
      "step": 500
    },
    {
      "epoch": 4.543429844097996,
      "grad_norm": 22.715709686279297,
      "learning_rate": 4.290674603174603e-05,
      "loss": 0.7042,
      "step": 510
    },
    {
      "epoch": 4.632516703786192,
      "grad_norm": 22.303823471069336,
      "learning_rate": 4.265873015873016e-05,
      "loss": 0.64,
      "step": 520
    },
    {
      "epoch": 4.721603563474387,
      "grad_norm": 24.049057006835938,
      "learning_rate": 4.2410714285714285e-05,
      "loss": 0.6607,
      "step": 530
    },
    {
      "epoch": 4.810690423162583,
      "grad_norm": 24.192787170410156,
      "learning_rate": 4.2162698412698416e-05,
      "loss": 0.6641,
      "step": 540
    },
    {
      "epoch": 4.8997772828507795,
      "grad_norm": 18.463960647583008,
      "learning_rate": 4.191468253968254e-05,
      "loss": 0.6725,
      "step": 550
    },
    {
      "epoch": 4.988864142538976,
      "grad_norm": 14.166847229003906,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.662,
      "step": 560
    },
    {
      "epoch": 4.997772828507795,
      "eval_accuracy": 0.817783343769568,
      "eval_loss": 0.5464382767677307,
      "eval_runtime": 21.2366,
      "eval_samples_per_second": 75.2,
      "eval_steps_per_second": 2.354,
      "step": 561
    },
    {
      "epoch": 5.077951002227172,
      "grad_norm": 17.516250610351562,
      "learning_rate": 4.14186507936508e-05,
      "loss": 0.5613,
      "step": 570
    },
    {
      "epoch": 5.167037861915367,
      "grad_norm": 16.576993942260742,
      "learning_rate": 4.117063492063492e-05,
      "loss": 0.6792,
      "step": 580
    },
    {
      "epoch": 5.256124721603563,
      "grad_norm": 21.692968368530273,
      "learning_rate": 4.0922619047619046e-05,
      "loss": 0.6623,
      "step": 590
    },
    {
      "epoch": 5.3452115812917596,
      "grad_norm": 19.450206756591797,
      "learning_rate": 4.067460317460318e-05,
      "loss": 0.5972,
      "step": 600
    },
    {
      "epoch": 5.434298440979956,
      "grad_norm": 15.104981422424316,
      "learning_rate": 4.04265873015873e-05,
      "loss": 0.5822,
      "step": 610
    },
    {
      "epoch": 5.523385300668151,
      "grad_norm": 18.1156005859375,
      "learning_rate": 4.017857142857143e-05,
      "loss": 0.5892,
      "step": 620
    },
    {
      "epoch": 5.612472160356347,
      "grad_norm": 17.83465576171875,
      "learning_rate": 3.993055555555556e-05,
      "loss": 0.5647,
      "step": 630
    },
    {
      "epoch": 5.701559020044543,
      "grad_norm": 21.32401466369629,
      "learning_rate": 3.968253968253968e-05,
      "loss": 0.6736,
      "step": 640
    },
    {
      "epoch": 5.79064587973274,
      "grad_norm": 20.872905731201172,
      "learning_rate": 3.943452380952381e-05,
      "loss": 0.6413,
      "step": 650
    },
    {
      "epoch": 5.879732739420936,
      "grad_norm": 18.514385223388672,
      "learning_rate": 3.918650793650794e-05,
      "loss": 0.5748,
      "step": 660
    },
    {
      "epoch": 5.968819599109131,
      "grad_norm": 15.119345664978027,
      "learning_rate": 3.893849206349206e-05,
      "loss": 0.6211,
      "step": 670
    },
    {
      "epoch": 5.99554565701559,
      "eval_accuracy": 0.8134001252348153,
      "eval_loss": 0.5666616559028625,
      "eval_runtime": 21.3956,
      "eval_samples_per_second": 74.642,
      "eval_steps_per_second": 2.337,
      "step": 673
    },
    {
      "epoch": 6.057906458797327,
      "grad_norm": 14.720823287963867,
      "learning_rate": 3.8690476190476195e-05,
      "loss": 0.5748,
      "step": 680
    },
    {
      "epoch": 6.146993318485523,
      "grad_norm": 18.489421844482422,
      "learning_rate": 3.844246031746032e-05,
      "loss": 0.5077,
      "step": 690
    },
    {
      "epoch": 6.23608017817372,
      "grad_norm": 20.35164451599121,
      "learning_rate": 3.8194444444444444e-05,
      "loss": 0.5337,
      "step": 700
    },
    {
      "epoch": 6.325167037861915,
      "grad_norm": 19.843276977539062,
      "learning_rate": 3.794642857142857e-05,
      "loss": 0.5381,
      "step": 710
    },
    {
      "epoch": 6.414253897550111,
      "grad_norm": 18.632654190063477,
      "learning_rate": 3.76984126984127e-05,
      "loss": 0.5889,
      "step": 720
    },
    {
      "epoch": 6.503340757238307,
      "grad_norm": 17.311256408691406,
      "learning_rate": 3.7450396825396824e-05,
      "loss": 0.5198,
      "step": 730
    },
    {
      "epoch": 6.5924276169265035,
      "grad_norm": 14.276023864746094,
      "learning_rate": 3.7202380952380956e-05,
      "loss": 0.5255,
      "step": 740
    },
    {
      "epoch": 6.6815144766147,
      "grad_norm": 18.370847702026367,
      "learning_rate": 3.695436507936508e-05,
      "loss": 0.5429,
      "step": 750
    },
    {
      "epoch": 6.770601336302895,
      "grad_norm": 15.048176765441895,
      "learning_rate": 3.6706349206349205e-05,
      "loss": 0.5933,
      "step": 760
    },
    {
      "epoch": 6.859688195991091,
      "grad_norm": 23.03609275817871,
      "learning_rate": 3.6458333333333336e-05,
      "loss": 0.5956,
      "step": 770
    },
    {
      "epoch": 6.948775055679287,
      "grad_norm": 16.367786407470703,
      "learning_rate": 3.621031746031746e-05,
      "loss": 0.5374,
      "step": 780
    },
    {
      "epoch": 6.993318485523385,
      "eval_accuracy": 0.8296806512210394,
      "eval_loss": 0.5134668946266174,
      "eval_runtime": 21.1708,
      "eval_samples_per_second": 75.434,
      "eval_steps_per_second": 2.362,
      "step": 785
    },
    {
      "epoch": 7.0378619153674835,
      "grad_norm": 14.078798294067383,
      "learning_rate": 3.5962301587301586e-05,
      "loss": 0.5302,
      "step": 790
    },
    {
      "epoch": 7.12694877505568,
      "grad_norm": 16.1302433013916,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.5558,
      "step": 800
    },
    {
      "epoch": 7.216035634743875,
      "grad_norm": 13.907476425170898,
      "learning_rate": 3.546626984126984e-05,
      "loss": 0.5582,
      "step": 810
    },
    {
      "epoch": 7.305122494432071,
      "grad_norm": 17.780229568481445,
      "learning_rate": 3.521825396825397e-05,
      "loss": 0.5502,
      "step": 820
    },
    {
      "epoch": 7.394209354120267,
      "grad_norm": 26.133169174194336,
      "learning_rate": 3.49702380952381e-05,
      "loss": 0.5396,
      "step": 830
    },
    {
      "epoch": 7.4832962138084635,
      "grad_norm": 13.426424980163574,
      "learning_rate": 3.472222222222222e-05,
      "loss": 0.5536,
      "step": 840
    },
    {
      "epoch": 7.57238307349666,
      "grad_norm": 24.7213077545166,
      "learning_rate": 3.4474206349206354e-05,
      "loss": 0.5081,
      "step": 850
    },
    {
      "epoch": 7.661469933184855,
      "grad_norm": 20.476024627685547,
      "learning_rate": 3.422619047619048e-05,
      "loss": 0.5186,
      "step": 860
    },
    {
      "epoch": 7.750556792873051,
      "grad_norm": 14.186267852783203,
      "learning_rate": 3.397817460317461e-05,
      "loss": 0.5517,
      "step": 870
    },
    {
      "epoch": 7.839643652561247,
      "grad_norm": 17.4648494720459,
      "learning_rate": 3.3730158730158734e-05,
      "loss": 0.546,
      "step": 880
    },
    {
      "epoch": 7.928730512249444,
      "grad_norm": 17.170608520507812,
      "learning_rate": 3.348214285714286e-05,
      "loss": 0.5147,
      "step": 890
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.828428303068253,
      "eval_loss": 0.5184165835380554,
      "eval_runtime": 21.3041,
      "eval_samples_per_second": 74.962,
      "eval_steps_per_second": 2.347,
      "step": 898
    },
    {
      "epoch": 8.017817371937639,
      "grad_norm": 15.437606811523438,
      "learning_rate": 3.3234126984126983e-05,
      "loss": 0.5755,
      "step": 900
    },
    {
      "epoch": 8.106904231625835,
      "grad_norm": 13.340350151062012,
      "learning_rate": 3.2986111111111115e-05,
      "loss": 0.4563,
      "step": 910
    },
    {
      "epoch": 8.195991091314031,
      "grad_norm": 11.855537414550781,
      "learning_rate": 3.273809523809524e-05,
      "loss": 0.4695,
      "step": 920
    },
    {
      "epoch": 8.285077951002227,
      "grad_norm": 19.414112091064453,
      "learning_rate": 3.249007936507937e-05,
      "loss": 0.4728,
      "step": 930
    },
    {
      "epoch": 8.374164810690424,
      "grad_norm": 14.449013710021973,
      "learning_rate": 3.2242063492063495e-05,
      "loss": 0.5105,
      "step": 940
    },
    {
      "epoch": 8.46325167037862,
      "grad_norm": 12.747797966003418,
      "learning_rate": 3.199404761904762e-05,
      "loss": 0.5009,
      "step": 950
    },
    {
      "epoch": 8.552338530066816,
      "grad_norm": 17.288352966308594,
      "learning_rate": 3.1746031746031745e-05,
      "loss": 0.5235,
      "step": 960
    },
    {
      "epoch": 8.64142538975501,
      "grad_norm": 10.824517250061035,
      "learning_rate": 3.1498015873015876e-05,
      "loss": 0.4365,
      "step": 970
    },
    {
      "epoch": 8.730512249443207,
      "grad_norm": 11.664321899414062,
      "learning_rate": 3.125e-05,
      "loss": 0.4819,
      "step": 980
    },
    {
      "epoch": 8.819599109131403,
      "grad_norm": 15.481782913208008,
      "learning_rate": 3.100198412698413e-05,
      "loss": 0.4624,
      "step": 990
    },
    {
      "epoch": 8.908685968819599,
      "grad_norm": 15.801815032958984,
      "learning_rate": 3.075396825396826e-05,
      "loss": 0.4689,
      "step": 1000
    },
    {
      "epoch": 8.997772828507795,
      "grad_norm": 12.902999877929688,
      "learning_rate": 3.0505952380952385e-05,
      "loss": 0.5101,
      "step": 1010
    },
    {
      "epoch": 8.997772828507795,
      "eval_accuracy": 0.8090169067000627,
      "eval_loss": 0.5775282382965088,
      "eval_runtime": 21.2996,
      "eval_samples_per_second": 74.978,
      "eval_steps_per_second": 2.347,
      "step": 1010
    },
    {
      "epoch": 9.086859688195991,
      "grad_norm": 14.9346342086792,
      "learning_rate": 3.0257936507936506e-05,
      "loss": 0.4874,
      "step": 1020
    },
    {
      "epoch": 9.175946547884188,
      "grad_norm": 13.989501953125,
      "learning_rate": 3.0009920634920634e-05,
      "loss": 0.4677,
      "step": 1030
    },
    {
      "epoch": 9.265033407572384,
      "grad_norm": 13.029789924621582,
      "learning_rate": 2.9761904761904762e-05,
      "loss": 0.4558,
      "step": 1040
    },
    {
      "epoch": 9.35412026726058,
      "grad_norm": 13.129937171936035,
      "learning_rate": 2.951388888888889e-05,
      "loss": 0.477,
      "step": 1050
    },
    {
      "epoch": 9.443207126948774,
      "grad_norm": 18.104257583618164,
      "learning_rate": 2.9265873015873018e-05,
      "loss": 0.4678,
      "step": 1060
    },
    {
      "epoch": 9.53229398663697,
      "grad_norm": 14.129535675048828,
      "learning_rate": 2.9017857142857146e-05,
      "loss": 0.5203,
      "step": 1070
    },
    {
      "epoch": 9.621380846325167,
      "grad_norm": 12.165337562561035,
      "learning_rate": 2.876984126984127e-05,
      "loss": 0.4555,
      "step": 1080
    },
    {
      "epoch": 9.710467706013363,
      "grad_norm": 11.243370056152344,
      "learning_rate": 2.8521825396825395e-05,
      "loss": 0.4713,
      "step": 1090
    },
    {
      "epoch": 9.799554565701559,
      "grad_norm": 13.38561725616455,
      "learning_rate": 2.8273809523809523e-05,
      "loss": 0.4461,
      "step": 1100
    },
    {
      "epoch": 9.888641425389755,
      "grad_norm": 24.981054306030273,
      "learning_rate": 2.802579365079365e-05,
      "loss": 0.4867,
      "step": 1110
    },
    {
      "epoch": 9.977728285077951,
      "grad_norm": 15.780611991882324,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.4791,
      "step": 1120
    },
    {
      "epoch": 9.99554565701559,
      "eval_accuracy": 0.8278021289918598,
      "eval_loss": 0.49649718403816223,
      "eval_runtime": 21.0785,
      "eval_samples_per_second": 75.765,
      "eval_steps_per_second": 2.372,
      "step": 1122
    },
    {
      "epoch": 10.066815144766148,
      "grad_norm": 20.694881439208984,
      "learning_rate": 2.7529761904761907e-05,
      "loss": 0.4615,
      "step": 1130
    },
    {
      "epoch": 10.155902004454344,
      "grad_norm": 7.557748317718506,
      "learning_rate": 2.7281746031746032e-05,
      "loss": 0.5191,
      "step": 1140
    },
    {
      "epoch": 10.244988864142538,
      "grad_norm": 21.552391052246094,
      "learning_rate": 2.703373015873016e-05,
      "loss": 0.4283,
      "step": 1150
    },
    {
      "epoch": 10.334075723830734,
      "grad_norm": 16.425872802734375,
      "learning_rate": 2.6785714285714288e-05,
      "loss": 0.4424,
      "step": 1160
    },
    {
      "epoch": 10.42316258351893,
      "grad_norm": 18.903982162475586,
      "learning_rate": 2.6537698412698416e-05,
      "loss": 0.4488,
      "step": 1170
    },
    {
      "epoch": 10.512249443207127,
      "grad_norm": 20.344375610351562,
      "learning_rate": 2.628968253968254e-05,
      "loss": 0.4788,
      "step": 1180
    },
    {
      "epoch": 10.601336302895323,
      "grad_norm": 9.902765274047852,
      "learning_rate": 2.604166666666667e-05,
      "loss": 0.4644,
      "step": 1190
    },
    {
      "epoch": 10.690423162583519,
      "grad_norm": 14.26260757446289,
      "learning_rate": 2.5793650793650796e-05,
      "loss": 0.3828,
      "step": 1200
    },
    {
      "epoch": 10.779510022271715,
      "grad_norm": 10.488040924072266,
      "learning_rate": 2.554563492063492e-05,
      "loss": 0.4348,
      "step": 1210
    },
    {
      "epoch": 10.868596881959911,
      "grad_norm": 19.645410537719727,
      "learning_rate": 2.529761904761905e-05,
      "loss": 0.4537,
      "step": 1220
    },
    {
      "epoch": 10.957683741648108,
      "grad_norm": 13.607442855834961,
      "learning_rate": 2.5049603174603177e-05,
      "loss": 0.4339,
      "step": 1230
    },
    {
      "epoch": 10.993318485523385,
      "eval_accuracy": 0.8303068252974327,
      "eval_loss": 0.5053770542144775,
      "eval_runtime": 21.1844,
      "eval_samples_per_second": 75.386,
      "eval_steps_per_second": 2.36,
      "step": 1234
    },
    {
      "epoch": 11.046770601336302,
      "grad_norm": 10.476160049438477,
      "learning_rate": 2.4801587301587305e-05,
      "loss": 0.3619,
      "step": 1240
    },
    {
      "epoch": 11.135857461024498,
      "grad_norm": 12.624418258666992,
      "learning_rate": 2.455357142857143e-05,
      "loss": 0.3775,
      "step": 1250
    },
    {
      "epoch": 11.224944320712694,
      "grad_norm": 14.461904525756836,
      "learning_rate": 2.4305555555555558e-05,
      "loss": 0.447,
      "step": 1260
    },
    {
      "epoch": 11.31403118040089,
      "grad_norm": 12.849555969238281,
      "learning_rate": 2.4057539682539686e-05,
      "loss": 0.3806,
      "step": 1270
    },
    {
      "epoch": 11.403118040089087,
      "grad_norm": 13.973437309265137,
      "learning_rate": 2.380952380952381e-05,
      "loss": 0.4338,
      "step": 1280
    },
    {
      "epoch": 11.492204899777283,
      "grad_norm": 9.459921836853027,
      "learning_rate": 2.3561507936507938e-05,
      "loss": 0.4273,
      "step": 1290
    },
    {
      "epoch": 11.58129175946548,
      "grad_norm": 19.047231674194336,
      "learning_rate": 2.3313492063492066e-05,
      "loss": 0.4625,
      "step": 1300
    },
    {
      "epoch": 11.670378619153675,
      "grad_norm": 14.69473934173584,
      "learning_rate": 2.3065476190476194e-05,
      "loss": 0.434,
      "step": 1310
    },
    {
      "epoch": 11.759465478841872,
      "grad_norm": 12.16936206817627,
      "learning_rate": 2.281746031746032e-05,
      "loss": 0.4532,
      "step": 1320
    },
    {
      "epoch": 11.848552338530066,
      "grad_norm": 11.761576652526855,
      "learning_rate": 2.2569444444444447e-05,
      "loss": 0.3755,
      "step": 1330
    },
    {
      "epoch": 11.937639198218262,
      "grad_norm": 15.081666946411133,
      "learning_rate": 2.2321428571428575e-05,
      "loss": 0.4709,
      "step": 1340
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8033813400125235,
      "eval_loss": 0.6274072527885437,
      "eval_runtime": 21.1787,
      "eval_samples_per_second": 75.406,
      "eval_steps_per_second": 2.361,
      "step": 1347
    },
    {
      "epoch": 12.026726057906458,
      "grad_norm": 15.724143028259277,
      "learning_rate": 2.20734126984127e-05,
      "loss": 0.4501,
      "step": 1350
    },
    {
      "epoch": 12.115812917594655,
      "grad_norm": 9.250158309936523,
      "learning_rate": 2.1825396825396827e-05,
      "loss": 0.3923,
      "step": 1360
    },
    {
      "epoch": 12.20489977728285,
      "grad_norm": 17.66366195678711,
      "learning_rate": 2.1577380952380955e-05,
      "loss": 0.3874,
      "step": 1370
    },
    {
      "epoch": 12.293986636971047,
      "grad_norm": 9.035892486572266,
      "learning_rate": 2.132936507936508e-05,
      "loss": 0.4155,
      "step": 1380
    },
    {
      "epoch": 12.383073496659243,
      "grad_norm": 14.44294548034668,
      "learning_rate": 2.1081349206349208e-05,
      "loss": 0.4047,
      "step": 1390
    },
    {
      "epoch": 12.47216035634744,
      "grad_norm": 12.312219619750977,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.4204,
      "step": 1400
    },
    {
      "epoch": 12.561247216035635,
      "grad_norm": 11.931431770324707,
      "learning_rate": 2.058531746031746e-05,
      "loss": 0.4006,
      "step": 1410
    },
    {
      "epoch": 12.65033407572383,
      "grad_norm": 9.212520599365234,
      "learning_rate": 2.033730158730159e-05,
      "loss": 0.3722,
      "step": 1420
    },
    {
      "epoch": 12.739420935412026,
      "grad_norm": 13.458575248718262,
      "learning_rate": 2.0089285714285717e-05,
      "loss": 0.3816,
      "step": 1430
    },
    {
      "epoch": 12.828507795100222,
      "grad_norm": 9.76841926574707,
      "learning_rate": 1.984126984126984e-05,
      "loss": 0.3839,
      "step": 1440
    },
    {
      "epoch": 12.917594654788418,
      "grad_norm": 15.495878219604492,
      "learning_rate": 1.959325396825397e-05,
      "loss": 0.4092,
      "step": 1450
    },
    {
      "epoch": 12.997772828507795,
      "eval_accuracy": 0.8484658735128365,
      "eval_loss": 0.47939950227737427,
      "eval_runtime": 21.0587,
      "eval_samples_per_second": 75.836,
      "eval_steps_per_second": 2.374,
      "step": 1459
    },
    {
      "epoch": 13.006681514476615,
      "grad_norm": 10.429738998413086,
      "learning_rate": 1.9345238095238097e-05,
      "loss": 0.3871,
      "step": 1460
    },
    {
      "epoch": 13.09576837416481,
      "grad_norm": 10.946513175964355,
      "learning_rate": 1.9097222222222222e-05,
      "loss": 0.348,
      "step": 1470
    },
    {
      "epoch": 13.184855233853007,
      "grad_norm": 9.279400825500488,
      "learning_rate": 1.884920634920635e-05,
      "loss": 0.3221,
      "step": 1480
    },
    {
      "epoch": 13.273942093541203,
      "grad_norm": 13.357484817504883,
      "learning_rate": 1.8601190476190478e-05,
      "loss": 0.3058,
      "step": 1490
    },
    {
      "epoch": 13.3630289532294,
      "grad_norm": 13.243049621582031,
      "learning_rate": 1.8353174603174602e-05,
      "loss": 0.3596,
      "step": 1500
    },
    {
      "epoch": 13.452115812917596,
      "grad_norm": 9.535561561584473,
      "learning_rate": 1.810515873015873e-05,
      "loss": 0.34,
      "step": 1510
    },
    {
      "epoch": 13.54120267260579,
      "grad_norm": 11.322809219360352,
      "learning_rate": 1.785714285714286e-05,
      "loss": 0.3468,
      "step": 1520
    },
    {
      "epoch": 13.630289532293986,
      "grad_norm": 13.22533130645752,
      "learning_rate": 1.7609126984126986e-05,
      "loss": 0.3749,
      "step": 1530
    },
    {
      "epoch": 13.719376391982182,
      "grad_norm": 9.667810440063477,
      "learning_rate": 1.736111111111111e-05,
      "loss": 0.3601,
      "step": 1540
    },
    {
      "epoch": 13.808463251670378,
      "grad_norm": 16.140737533569336,
      "learning_rate": 1.711309523809524e-05,
      "loss": 0.3634,
      "step": 1550
    },
    {
      "epoch": 13.897550111358575,
      "grad_norm": 13.473885536193848,
      "learning_rate": 1.6865079365079367e-05,
      "loss": 0.3983,
      "step": 1560
    },
    {
      "epoch": 13.98663697104677,
      "grad_norm": 13.371350288391113,
      "learning_rate": 1.6617063492063492e-05,
      "loss": 0.368,
      "step": 1570
    },
    {
      "epoch": 13.99554565701559,
      "eval_accuracy": 0.8465873512836568,
      "eval_loss": 0.5031822919845581,
      "eval_runtime": 21.1723,
      "eval_samples_per_second": 75.429,
      "eval_steps_per_second": 2.362,
      "step": 1571
    },
    {
      "epoch": 14.075723830734967,
      "grad_norm": 11.687047958374023,
      "learning_rate": 1.636904761904762e-05,
      "loss": 0.3465,
      "step": 1580
    },
    {
      "epoch": 14.164810690423163,
      "grad_norm": 13.193872451782227,
      "learning_rate": 1.6121031746031748e-05,
      "loss": 0.3402,
      "step": 1590
    },
    {
      "epoch": 14.25389755011136,
      "grad_norm": 15.279977798461914,
      "learning_rate": 1.5873015873015872e-05,
      "loss": 0.358,
      "step": 1600
    },
    {
      "epoch": 14.342984409799554,
      "grad_norm": 10.373139381408691,
      "learning_rate": 1.5625e-05,
      "loss": 0.3423,
      "step": 1610
    },
    {
      "epoch": 14.43207126948775,
      "grad_norm": 14.441068649291992,
      "learning_rate": 1.537698412698413e-05,
      "loss": 0.323,
      "step": 1620
    },
    {
      "epoch": 14.521158129175946,
      "grad_norm": 13.793278694152832,
      "learning_rate": 1.5128968253968253e-05,
      "loss": 0.3443,
      "step": 1630
    },
    {
      "epoch": 14.610244988864142,
      "grad_norm": 13.510929107666016,
      "learning_rate": 1.4880952380952381e-05,
      "loss": 0.3289,
      "step": 1640
    },
    {
      "epoch": 14.699331848552339,
      "grad_norm": 14.0494384765625,
      "learning_rate": 1.4632936507936509e-05,
      "loss": 0.3266,
      "step": 1650
    },
    {
      "epoch": 14.788418708240535,
      "grad_norm": 17.658891677856445,
      "learning_rate": 1.4384920634920635e-05,
      "loss": 0.343,
      "step": 1660
    },
    {
      "epoch": 14.877505567928731,
      "grad_norm": 14.786270141601562,
      "learning_rate": 1.4136904761904762e-05,
      "loss": 0.3262,
      "step": 1670
    },
    {
      "epoch": 14.966592427616927,
      "grad_norm": 11.525693893432617,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.3415,
      "step": 1680
    },
    {
      "epoch": 14.993318485523385,
      "eval_accuracy": 0.8409517845961177,
      "eval_loss": 0.5257949829101562,
      "eval_runtime": 21.2165,
      "eval_samples_per_second": 75.272,
      "eval_steps_per_second": 2.357,
      "step": 1683
    },
    {
      "epoch": 15.055679287305123,
      "grad_norm": 12.09720516204834,
      "learning_rate": 1.3640873015873016e-05,
      "loss": 0.276,
      "step": 1690
    },
    {
      "epoch": 15.144766146993318,
      "grad_norm": 15.346843719482422,
      "learning_rate": 1.3392857142857144e-05,
      "loss": 0.3415,
      "step": 1700
    },
    {
      "epoch": 15.233853006681514,
      "grad_norm": 10.313881874084473,
      "learning_rate": 1.314484126984127e-05,
      "loss": 0.2949,
      "step": 1710
    },
    {
      "epoch": 15.32293986636971,
      "grad_norm": 11.033019065856934,
      "learning_rate": 1.2896825396825398e-05,
      "loss": 0.3653,
      "step": 1720
    },
    {
      "epoch": 15.412026726057906,
      "grad_norm": 12.691286087036133,
      "learning_rate": 1.2648809523809524e-05,
      "loss": 0.3321,
      "step": 1730
    },
    {
      "epoch": 15.501113585746102,
      "grad_norm": 8.7056884765625,
      "learning_rate": 1.2400793650793652e-05,
      "loss": 0.3351,
      "step": 1740
    },
    {
      "epoch": 15.590200445434299,
      "grad_norm": 17.608694076538086,
      "learning_rate": 1.2152777777777779e-05,
      "loss": 0.2772,
      "step": 1750
    },
    {
      "epoch": 15.679287305122495,
      "grad_norm": 12.325498580932617,
      "learning_rate": 1.1904761904761905e-05,
      "loss": 0.3475,
      "step": 1760
    },
    {
      "epoch": 15.768374164810691,
      "grad_norm": 13.088592529296875,
      "learning_rate": 1.1656746031746033e-05,
      "loss": 0.3217,
      "step": 1770
    },
    {
      "epoch": 15.857461024498887,
      "grad_norm": 10.744033813476562,
      "learning_rate": 1.140873015873016e-05,
      "loss": 0.3311,
      "step": 1780
    },
    {
      "epoch": 15.946547884187082,
      "grad_norm": 12.45305347442627,
      "learning_rate": 1.1160714285714287e-05,
      "loss": 0.2963,
      "step": 1790
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.841577958672511,
      "eval_loss": 0.5382400155067444,
      "eval_runtime": 21.0587,
      "eval_samples_per_second": 75.836,
      "eval_steps_per_second": 2.374,
      "step": 1796
    },
    {
      "epoch": 16.035634743875278,
      "grad_norm": 13.57647705078125,
      "learning_rate": 1.0912698412698414e-05,
      "loss": 0.3066,
      "step": 1800
    },
    {
      "epoch": 16.124721603563476,
      "grad_norm": 13.313094139099121,
      "learning_rate": 1.066468253968254e-05,
      "loss": 0.2948,
      "step": 1810
    },
    {
      "epoch": 16.21380846325167,
      "grad_norm": 10.375079154968262,
      "learning_rate": 1.0416666666666668e-05,
      "loss": 0.2852,
      "step": 1820
    },
    {
      "epoch": 16.302895322939868,
      "grad_norm": 10.751969337463379,
      "learning_rate": 1.0168650793650794e-05,
      "loss": 0.3016,
      "step": 1830
    },
    {
      "epoch": 16.391982182628063,
      "grad_norm": 9.340880393981934,
      "learning_rate": 9.92063492063492e-06,
      "loss": 0.2849,
      "step": 1840
    },
    {
      "epoch": 16.481069042316257,
      "grad_norm": 11.936270713806152,
      "learning_rate": 9.672619047619049e-06,
      "loss": 0.2715,
      "step": 1850
    },
    {
      "epoch": 16.570155902004455,
      "grad_norm": 12.299432754516602,
      "learning_rate": 9.424603174603175e-06,
      "loss": 0.2872,
      "step": 1860
    },
    {
      "epoch": 16.65924276169265,
      "grad_norm": 11.702118873596191,
      "learning_rate": 9.176587301587301e-06,
      "loss": 0.2865,
      "step": 1870
    },
    {
      "epoch": 16.748329621380847,
      "grad_norm": 9.740532875061035,
      "learning_rate": 8.92857142857143e-06,
      "loss": 0.2727,
      "step": 1880
    },
    {
      "epoch": 16.83741648106904,
      "grad_norm": 12.387954711914062,
      "learning_rate": 8.680555555555556e-06,
      "loss": 0.3054,
      "step": 1890
    },
    {
      "epoch": 16.92650334075724,
      "grad_norm": 17.06888198852539,
      "learning_rate": 8.432539682539684e-06,
      "loss": 0.2614,
      "step": 1900
    },
    {
      "epoch": 16.997772828507795,
      "eval_accuracy": 0.8559799624295554,
      "eval_loss": 0.5104979872703552,
      "eval_runtime": 21.2594,
      "eval_samples_per_second": 75.12,
      "eval_steps_per_second": 2.352,
      "step": 1908
    },
    {
      "epoch": 17.015590200445434,
      "grad_norm": 9.6270170211792,
      "learning_rate": 8.18452380952381e-06,
      "loss": 0.3009,
      "step": 1910
    },
    {
      "epoch": 17.104677060133632,
      "grad_norm": 14.768975257873535,
      "learning_rate": 7.936507936507936e-06,
      "loss": 0.2854,
      "step": 1920
    },
    {
      "epoch": 17.193763919821826,
      "grad_norm": 15.029583930969238,
      "learning_rate": 7.688492063492064e-06,
      "loss": 0.2663,
      "step": 1930
    },
    {
      "epoch": 17.28285077951002,
      "grad_norm": 6.971687316894531,
      "learning_rate": 7.4404761904761905e-06,
      "loss": 0.2263,
      "step": 1940
    },
    {
      "epoch": 17.37193763919822,
      "grad_norm": 12.997879981994629,
      "learning_rate": 7.192460317460318e-06,
      "loss": 0.2366,
      "step": 1950
    },
    {
      "epoch": 17.461024498886413,
      "grad_norm": 20.721641540527344,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.2708,
      "step": 1960
    },
    {
      "epoch": 17.55011135857461,
      "grad_norm": 8.18585205078125,
      "learning_rate": 6.696428571428572e-06,
      "loss": 0.2462,
      "step": 1970
    },
    {
      "epoch": 17.639198218262806,
      "grad_norm": 9.622570991516113,
      "learning_rate": 6.448412698412699e-06,
      "loss": 0.282,
      "step": 1980
    },
    {
      "epoch": 17.728285077951004,
      "grad_norm": 9.59653377532959,
      "learning_rate": 6.200396825396826e-06,
      "loss": 0.2607,
      "step": 1990
    },
    {
      "epoch": 17.817371937639198,
      "grad_norm": 9.513813972473145,
      "learning_rate": 5.9523809523809525e-06,
      "loss": 0.2732,
      "step": 2000
    },
    {
      "epoch": 17.906458797327396,
      "grad_norm": 11.091604232788086,
      "learning_rate": 5.70436507936508e-06,
      "loss": 0.2512,
      "step": 2010
    },
    {
      "epoch": 17.99554565701559,
      "grad_norm": 11.867079734802246,
      "learning_rate": 5.456349206349207e-06,
      "loss": 0.2542,
      "step": 2020
    },
    {
      "epoch": 17.99554565701559,
      "eval_accuracy": 0.8534752661239825,
      "eval_loss": 0.5331311821937561,
      "eval_runtime": 21.0758,
      "eval_samples_per_second": 75.774,
      "eval_steps_per_second": 2.372,
      "step": 2020
    },
    {
      "epoch": 18.084632516703785,
      "grad_norm": 19.73405647277832,
      "learning_rate": 5.208333333333334e-06,
      "loss": 0.2383,
      "step": 2030
    },
    {
      "epoch": 18.173719376391983,
      "grad_norm": 9.715604782104492,
      "learning_rate": 4.96031746031746e-06,
      "loss": 0.2284,
      "step": 2040
    },
    {
      "epoch": 18.262806236080177,
      "grad_norm": 10.271077156066895,
      "learning_rate": 4.7123015873015875e-06,
      "loss": 0.2181,
      "step": 2050
    },
    {
      "epoch": 18.351893095768375,
      "grad_norm": 14.901619911193848,
      "learning_rate": 4.464285714285715e-06,
      "loss": 0.2642,
      "step": 2060
    },
    {
      "epoch": 18.44097995545657,
      "grad_norm": 8.47389030456543,
      "learning_rate": 4.216269841269842e-06,
      "loss": 0.2703,
      "step": 2070
    },
    {
      "epoch": 18.530066815144767,
      "grad_norm": 18.4124755859375,
      "learning_rate": 3.968253968253968e-06,
      "loss": 0.2427,
      "step": 2080
    },
    {
      "epoch": 18.619153674832962,
      "grad_norm": 6.289044380187988,
      "learning_rate": 3.7202380952380952e-06,
      "loss": 0.2784,
      "step": 2090
    },
    {
      "epoch": 18.70824053452116,
      "grad_norm": 8.155685424804688,
      "learning_rate": 3.4722222222222224e-06,
      "loss": 0.2096,
      "step": 2100
    },
    {
      "epoch": 18.797327394209354,
      "grad_norm": 10.127546310424805,
      "learning_rate": 3.2242063492063495e-06,
      "loss": 0.2247,
      "step": 2110
    },
    {
      "epoch": 18.88641425389755,
      "grad_norm": 8.079633712768555,
      "learning_rate": 2.9761904761904763e-06,
      "loss": 0.2291,
      "step": 2120
    },
    {
      "epoch": 18.975501113585747,
      "grad_norm": 11.977538108825684,
      "learning_rate": 2.7281746031746034e-06,
      "loss": 0.2436,
      "step": 2130
    },
    {
      "epoch": 18.993318485523385,
      "eval_accuracy": 0.8609893550407013,
      "eval_loss": 0.5468289256095886,
      "eval_runtime": 21.1931,
      "eval_samples_per_second": 75.355,
      "eval_steps_per_second": 2.359,
      "step": 2132
    },
    {
      "epoch": 19.06458797327394,
      "grad_norm": 11.12005615234375,
      "learning_rate": 2.48015873015873e-06,
      "loss": 0.2046,
      "step": 2140
    },
    {
      "epoch": 19.15367483296214,
      "grad_norm": 8.387917518615723,
      "learning_rate": 2.2321428571428573e-06,
      "loss": 0.2178,
      "step": 2150
    },
    {
      "epoch": 19.242761692650333,
      "grad_norm": 14.341036796569824,
      "learning_rate": 1.984126984126984e-06,
      "loss": 0.2194,
      "step": 2160
    },
    {
      "epoch": 19.33184855233853,
      "grad_norm": 16.983957290649414,
      "learning_rate": 1.7361111111111112e-06,
      "loss": 0.2644,
      "step": 2170
    },
    {
      "epoch": 19.420935412026726,
      "grad_norm": 6.247664928436279,
      "learning_rate": 1.4880952380952381e-06,
      "loss": 0.2198,
      "step": 2180
    },
    {
      "epoch": 19.510022271714924,
      "grad_norm": 15.509531021118164,
      "learning_rate": 1.240079365079365e-06,
      "loss": 0.2524,
      "step": 2190
    },
    {
      "epoch": 19.599109131403118,
      "grad_norm": 8.883134841918945,
      "learning_rate": 9.92063492063492e-07,
      "loss": 0.2226,
      "step": 2200
    },
    {
      "epoch": 19.688195991091312,
      "grad_norm": 8.381498336791992,
      "learning_rate": 7.440476190476191e-07,
      "loss": 0.2067,
      "step": 2210
    },
    {
      "epoch": 19.77728285077951,
      "grad_norm": 9.45773983001709,
      "learning_rate": 4.96031746031746e-07,
      "loss": 0.2116,
      "step": 2220
    },
    {
      "epoch": 19.866369710467705,
      "grad_norm": 11.97305965423584,
      "learning_rate": 2.48015873015873e-07,
      "loss": 0.212,
      "step": 2230
    },
    {
      "epoch": 19.955456570155903,
      "grad_norm": 7.919186592102051,
      "learning_rate": 0.0,
      "loss": 0.2275,
      "step": 2240
    },
    {
      "epoch": 19.955456570155903,
      "eval_accuracy": 0.8703819661865999,
      "eval_loss": 0.5583016276359558,
      "eval_runtime": 21.2008,
      "eval_samples_per_second": 75.327,
      "eval_steps_per_second": 2.358,
      "step": 2240
    },
    {
      "epoch": 19.955456570155903,
      "step": 2240,
      "total_flos": 2.928337686848471e+19,
      "train_loss": 0.5139281852969102,
      "train_runtime": 9596.9456,
      "train_samples_per_second": 29.937,
      "train_steps_per_second": 0.233
    }
  ],
  "logging_steps": 10,
  "max_steps": 2240,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 2.928337686848471e+19,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}