{
  "best_metric": 0.5918110609054565,
  "best_model_checkpoint": "./kd_results/microsoft/beit-base-patch16-224_alpha0.5_temp5.0_t3/checkpoint-2624",
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 3280,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1524390243902439,
      "grad_norm": 4.68762731552124,
      "learning_rate": 3.8109756097560976e-06,
      "loss": 0.8428,
      "step": 25
    },
    {
      "epoch": 0.3048780487804878,
      "grad_norm": 5.215528964996338,
      "learning_rate": 7.621951219512195e-06,
      "loss": 0.8225,
      "step": 50
    },
    {
      "epoch": 0.4573170731707317,
      "grad_norm": 6.478719711303711,
      "learning_rate": 1.1432926829268294e-05,
      "loss": 0.8244,
      "step": 75
    },
    {
      "epoch": 0.6097560975609756,
      "grad_norm": 3.6382999420166016,
      "learning_rate": 1.524390243902439e-05,
      "loss": 0.8091,
      "step": 100
    },
    {
      "epoch": 0.7621951219512195,
      "grad_norm": 3.260287046432495,
      "learning_rate": 1.9054878048780488e-05,
      "loss": 0.8145,
      "step": 125
    },
    {
      "epoch": 0.9146341463414634,
      "grad_norm": 5.04181432723999,
      "learning_rate": 2.286585365853659e-05,
      "loss": 0.8134,
      "step": 150
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.25488069414316705,
      "eval_loss": 1.4050709009170532,
      "eval_runtime": 41.1635,
      "eval_samples_per_second": 22.398,
      "eval_steps_per_second": 0.705,
      "step": 164
    },
    {
      "epoch": 1.0670731707317074,
      "grad_norm": 3.7207865715026855,
      "learning_rate": 2.6676829268292686e-05,
      "loss": 0.8206,
      "step": 175
    },
    {
      "epoch": 1.2195121951219512,
      "grad_norm": 4.403680801391602,
      "learning_rate": 3.048780487804878e-05,
      "loss": 0.8063,
      "step": 200
    },
    {
      "epoch": 1.3719512195121952,
      "grad_norm": 6.912205219268799,
      "learning_rate": 3.429878048780488e-05,
      "loss": 0.7995,
      "step": 225
    },
    {
      "epoch": 1.524390243902439,
      "grad_norm": 2.571458339691162,
      "learning_rate": 3.8109756097560976e-05,
      "loss": 0.8111,
      "step": 250
    },
    {
      "epoch": 1.6768292682926829,
      "grad_norm": 4.953147888183594,
      "learning_rate": 4.1920731707317077e-05,
      "loss": 0.7909,
      "step": 275
    },
    {
      "epoch": 1.8292682926829267,
      "grad_norm": 3.780155658721924,
      "learning_rate": 4.573170731707318e-05,
      "loss": 0.7532,
      "step": 300
    },
    {
      "epoch": 1.9817073170731707,
      "grad_norm": 2.8478639125823975,
      "learning_rate": 4.954268292682927e-05,
      "loss": 0.7231,
      "step": 325
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.49783080260303686,
      "eval_loss": 1.1579526662826538,
      "eval_runtime": 39.9737,
      "eval_samples_per_second": 23.065,
      "eval_steps_per_second": 0.725,
      "step": 328
    },
    {
      "epoch": 2.1341463414634148,
      "grad_norm": 2.9338808059692383,
      "learning_rate": 4.962737127371274e-05,
      "loss": 0.6472,
      "step": 350
    },
    {
      "epoch": 2.2865853658536586,
      "grad_norm": 4.0507354736328125,
      "learning_rate": 4.92039295392954e-05,
      "loss": 0.6182,
      "step": 375
    },
    {
      "epoch": 2.4390243902439024,
      "grad_norm": 5.048892498016357,
      "learning_rate": 4.878048780487805e-05,
      "loss": 0.591,
      "step": 400
    },
    {
      "epoch": 2.591463414634146,
      "grad_norm": 7.343005180358887,
      "learning_rate": 4.8357046070460704e-05,
      "loss": 0.5737,
      "step": 425
    },
    {
      "epoch": 2.7439024390243905,
      "grad_norm": 5.656530380249023,
      "learning_rate": 4.793360433604336e-05,
      "loss": 0.5685,
      "step": 450
    },
    {
      "epoch": 2.8963414634146343,
      "grad_norm": 4.829864501953125,
      "learning_rate": 4.7510162601626016e-05,
      "loss": 0.538,
      "step": 475
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7581344902386117,
      "eval_loss": 0.7405389547348022,
      "eval_runtime": 42.2463,
      "eval_samples_per_second": 21.824,
      "eval_steps_per_second": 0.686,
      "step": 492
    },
    {
      "epoch": 3.048780487804878,
      "grad_norm": 3.7712740898132324,
      "learning_rate": 4.7086720867208675e-05,
      "loss": 0.5059,
      "step": 500
    },
    {
      "epoch": 3.201219512195122,
      "grad_norm": 2.8125956058502197,
      "learning_rate": 4.6663279132791334e-05,
      "loss": 0.4592,
      "step": 525
    },
    {
      "epoch": 3.3536585365853657,
      "grad_norm": 2.0605647563934326,
      "learning_rate": 4.6239837398373986e-05,
      "loss": 0.4445,
      "step": 550
    },
    {
      "epoch": 3.5060975609756095,
      "grad_norm": 3.0619406700134277,
      "learning_rate": 4.581639566395664e-05,
      "loss": 0.4696,
      "step": 575
    },
    {
      "epoch": 3.658536585365854,
      "grad_norm": 3.0977537631988525,
      "learning_rate": 4.53929539295393e-05,
      "loss": 0.4192,
      "step": 600
    },
    {
      "epoch": 3.8109756097560976,
      "grad_norm": 3.48732590675354,
      "learning_rate": 4.496951219512196e-05,
      "loss": 0.4256,
      "step": 625
    },
    {
      "epoch": 3.9634146341463414,
      "grad_norm": 4.01300573348999,
      "learning_rate": 4.454607046070461e-05,
      "loss": 0.4444,
      "step": 650
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7765726681127982,
      "eval_loss": 0.7383840084075928,
      "eval_runtime": 40.4228,
      "eval_samples_per_second": 22.809,
      "eval_steps_per_second": 0.717,
      "step": 656
    },
    {
      "epoch": 4.115853658536586,
      "grad_norm": 3.8454947471618652,
      "learning_rate": 4.412262872628726e-05,
      "loss": 0.4094,
      "step": 675
    },
    {
      "epoch": 4.2682926829268295,
      "grad_norm": 2.65144944190979,
      "learning_rate": 4.369918699186992e-05,
      "loss": 0.3767,
      "step": 700
    },
    {
      "epoch": 4.420731707317073,
      "grad_norm": 3.4917783737182617,
      "learning_rate": 4.3275745257452574e-05,
      "loss": 0.3798,
      "step": 725
    },
    {
      "epoch": 4.573170731707317,
      "grad_norm": 1.4538134336471558,
      "learning_rate": 4.285230352303523e-05,
      "loss": 0.3658,
      "step": 750
    },
    {
      "epoch": 4.725609756097561,
      "grad_norm": 3.654711961746216,
      "learning_rate": 4.242886178861789e-05,
      "loss": 0.3685,
      "step": 775
    },
    {
      "epoch": 4.878048780487805,
      "grad_norm": 2.776951313018799,
      "learning_rate": 4.2005420054200545e-05,
      "loss": 0.3712,
      "step": 800
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7809110629067245,
      "eval_loss": 0.6911107897758484,
      "eval_runtime": 39.5205,
      "eval_samples_per_second": 23.33,
      "eval_steps_per_second": 0.734,
      "step": 820
    },
    {
      "epoch": 5.030487804878049,
      "grad_norm": 2.283203601837158,
      "learning_rate": 4.15819783197832e-05,
      "loss": 0.3702,
      "step": 825
    },
    {
      "epoch": 5.182926829268292,
      "grad_norm": 1.1880710124969482,
      "learning_rate": 4.1158536585365856e-05,
      "loss": 0.3393,
      "step": 850
    },
    {
      "epoch": 5.335365853658536,
      "grad_norm": 4.259084701538086,
      "learning_rate": 4.0735094850948516e-05,
      "loss": 0.327,
      "step": 875
    },
    {
      "epoch": 5.487804878048781,
      "grad_norm": 2.179425001144409,
      "learning_rate": 4.031165311653117e-05,
      "loss": 0.3464,
      "step": 900
    },
    {
      "epoch": 5.640243902439025,
      "grad_norm": 2.059110403060913,
      "learning_rate": 3.988821138211382e-05,
      "loss": 0.3359,
      "step": 925
    },
    {
      "epoch": 5.7926829268292686,
      "grad_norm": 5.261742115020752,
      "learning_rate": 3.946476964769648e-05,
      "loss": 0.342,
      "step": 950
    },
    {
      "epoch": 5.945121951219512,
      "grad_norm": 2.6818530559539795,
      "learning_rate": 3.904132791327913e-05,
      "loss": 0.3419,
      "step": 975
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7939262472885033,
      "eval_loss": 0.6714470386505127,
      "eval_runtime": 40.826,
      "eval_samples_per_second": 22.584,
      "eval_steps_per_second": 0.71,
      "step": 984
    },
    {
      "epoch": 6.097560975609756,
      "grad_norm": 1.9220590591430664,
      "learning_rate": 3.861788617886179e-05,
      "loss": 0.3311,
      "step": 1000
    },
    {
      "epoch": 6.25,
      "grad_norm": 2.031047821044922,
      "learning_rate": 3.8194444444444444e-05,
      "loss": 0.3252,
      "step": 1025
    },
    {
      "epoch": 6.402439024390244,
      "grad_norm": 2.15692138671875,
      "learning_rate": 3.77710027100271e-05,
      "loss": 0.3213,
      "step": 1050
    },
    {
      "epoch": 6.554878048780488,
      "grad_norm": 1.0440400838851929,
      "learning_rate": 3.7347560975609755e-05,
      "loss": 0.3259,
      "step": 1075
    },
    {
      "epoch": 6.7073170731707314,
      "grad_norm": 2.292208671569824,
      "learning_rate": 3.6924119241192415e-05,
      "loss": 0.3185,
      "step": 1100
    },
    {
      "epoch": 6.859756097560975,
      "grad_norm": 1.1084694862365723,
      "learning_rate": 3.6500677506775074e-05,
      "loss": 0.3274,
      "step": 1125
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8167028199566161,
      "eval_loss": 0.6136873364448547,
      "eval_runtime": 41.2641,
      "eval_samples_per_second": 22.344,
      "eval_steps_per_second": 0.703,
      "step": 1148
    },
    {
      "epoch": 7.012195121951219,
      "grad_norm": 1.1337965726852417,
      "learning_rate": 3.6077235772357726e-05,
      "loss": 0.3067,
      "step": 1150
    },
    {
      "epoch": 7.164634146341464,
      "grad_norm": 3.3528525829315186,
      "learning_rate": 3.565379403794038e-05,
      "loss": 0.3105,
      "step": 1175
    },
    {
      "epoch": 7.317073170731708,
      "grad_norm": 2.85194730758667,
      "learning_rate": 3.523035230352303e-05,
      "loss": 0.3166,
      "step": 1200
    },
    {
      "epoch": 7.469512195121951,
      "grad_norm": 1.335073471069336,
      "learning_rate": 3.48069105691057e-05,
      "loss": 0.2998,
      "step": 1225
    },
    {
      "epoch": 7.621951219512195,
      "grad_norm": 0.8147186040878296,
      "learning_rate": 3.438346883468835e-05,
      "loss": 0.3002,
      "step": 1250
    },
    {
      "epoch": 7.774390243902439,
      "grad_norm": 3.3945751190185547,
      "learning_rate": 3.3960027100271e-05,
      "loss": 0.3118,
      "step": 1275
    },
    {
      "epoch": 7.926829268292683,
      "grad_norm": 1.6584522724151611,
      "learning_rate": 3.353658536585366e-05,
      "loss": 0.3137,
      "step": 1300
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8091106290672451,
      "eval_loss": 0.648438036441803,
      "eval_runtime": 40.4993,
      "eval_samples_per_second": 22.766,
      "eval_steps_per_second": 0.716,
      "step": 1312
    },
    {
      "epoch": 8.079268292682928,
      "grad_norm": 1.452438235282898,
      "learning_rate": 3.3113143631436314e-05,
      "loss": 0.3032,
      "step": 1325
    },
    {
      "epoch": 8.231707317073171,
      "grad_norm": 2.6878163814544678,
      "learning_rate": 3.268970189701897e-05,
      "loss": 0.2958,
      "step": 1350
    },
    {
      "epoch": 8.384146341463415,
      "grad_norm": 0.7613250613212585,
      "learning_rate": 3.226626016260163e-05,
      "loss": 0.3123,
      "step": 1375
    },
    {
      "epoch": 8.536585365853659,
      "grad_norm": 1.850319504737854,
      "learning_rate": 3.1842818428184285e-05,
      "loss": 0.3028,
      "step": 1400
    },
    {
      "epoch": 8.689024390243903,
      "grad_norm": 0.6270933151245117,
      "learning_rate": 3.141937669376694e-05,
      "loss": 0.2999,
      "step": 1425
    },
    {
      "epoch": 8.841463414634147,
      "grad_norm": 0.9790864586830139,
      "learning_rate": 3.0995934959349596e-05,
      "loss": 0.2988,
      "step": 1450
    },
    {
      "epoch": 8.99390243902439,
      "grad_norm": 1.1735106706619263,
      "learning_rate": 3.0572493224932255e-05,
      "loss": 0.3083,
      "step": 1475
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8156182212581344,
      "eval_loss": 0.6517806053161621,
      "eval_runtime": 40.9544,
      "eval_samples_per_second": 22.513,
      "eval_steps_per_second": 0.708,
      "step": 1476
    },
    {
      "epoch": 9.146341463414634,
      "grad_norm": 0.7330197095870972,
      "learning_rate": 3.0149051490514908e-05,
      "loss": 0.2924,
      "step": 1500
    },
    {
      "epoch": 9.298780487804878,
      "grad_norm": 1.7333564758300781,
      "learning_rate": 2.972560975609756e-05,
      "loss": 0.307,
      "step": 1525
    },
    {
      "epoch": 9.451219512195122,
      "grad_norm": 1.8797653913497925,
      "learning_rate": 2.930216802168022e-05,
      "loss": 0.2975,
      "step": 1550
    },
    {
      "epoch": 9.603658536585366,
      "grad_norm": 0.631047785282135,
      "learning_rate": 2.8878726287262875e-05,
      "loss": 0.3005,
      "step": 1575
    },
    {
      "epoch": 9.75609756097561,
      "grad_norm": 0.8189586400985718,
      "learning_rate": 2.8455284552845528e-05,
      "loss": 0.2983,
      "step": 1600
    },
    {
      "epoch": 9.908536585365853,
      "grad_norm": 1.3762580156326294,
      "learning_rate": 2.8031842818428184e-05,
      "loss": 0.2918,
      "step": 1625
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8101952277657267,
      "eval_loss": 0.6408398151397705,
      "eval_runtime": 40.7642,
      "eval_samples_per_second": 22.618,
      "eval_steps_per_second": 0.711,
      "step": 1640
    },
    {
      "epoch": 10.060975609756097,
      "grad_norm": 0.9083566665649414,
      "learning_rate": 2.7608401084010843e-05,
      "loss": 0.292,
      "step": 1650
    },
    {
      "epoch": 10.213414634146341,
      "grad_norm": 0.7562841176986694,
      "learning_rate": 2.71849593495935e-05,
      "loss": 0.2925,
      "step": 1675
    },
    {
      "epoch": 10.365853658536585,
      "grad_norm": 0.8686760067939758,
      "learning_rate": 2.676151761517615e-05,
      "loss": 0.2985,
      "step": 1700
    },
    {
      "epoch": 10.518292682926829,
      "grad_norm": 0.6039935350418091,
      "learning_rate": 2.633807588075881e-05,
      "loss": 0.2951,
      "step": 1725
    },
    {
      "epoch": 10.670731707317072,
      "grad_norm": 0.5477403402328491,
      "learning_rate": 2.5914634146341466e-05,
      "loss": 0.2901,
      "step": 1750
    },
    {
      "epoch": 10.823170731707316,
      "grad_norm": 0.701054573059082,
      "learning_rate": 2.549119241192412e-05,
      "loss": 0.295,
      "step": 1775
    },
    {
      "epoch": 10.975609756097562,
      "grad_norm": 1.4410309791564941,
      "learning_rate": 2.5067750677506774e-05,
      "loss": 0.2918,
      "step": 1800
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.8308026030368764,
      "eval_loss": 0.6469808220863342,
      "eval_runtime": 39.8921,
      "eval_samples_per_second": 23.112,
      "eval_steps_per_second": 0.727,
      "step": 1804
    },
    {
      "epoch": 11.128048780487806,
      "grad_norm": 0.9098992943763733,
      "learning_rate": 2.464430894308943e-05,
      "loss": 0.2916,
      "step": 1825
    },
    {
      "epoch": 11.28048780487805,
      "grad_norm": 0.7841638922691345,
      "learning_rate": 2.422086720867209e-05,
      "loss": 0.2854,
      "step": 1850
    },
    {
      "epoch": 11.432926829268293,
      "grad_norm": 0.6141788959503174,
      "learning_rate": 2.3797425474254745e-05,
      "loss": 0.2881,
      "step": 1875
    },
    {
      "epoch": 11.585365853658537,
      "grad_norm": 0.46465593576431274,
      "learning_rate": 2.3373983739837398e-05,
      "loss": 0.288,
      "step": 1900
    },
    {
      "epoch": 11.737804878048781,
      "grad_norm": 1.7072104215621948,
      "learning_rate": 2.2950542005420057e-05,
      "loss": 0.2996,
      "step": 1925
    },
    {
      "epoch": 11.890243902439025,
      "grad_norm": 0.6564107537269592,
      "learning_rate": 2.252710027100271e-05,
      "loss": 0.2915,
      "step": 1950
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8297180043383948,
      "eval_loss": 0.6152055263519287,
      "eval_runtime": 41.21,
      "eval_samples_per_second": 22.373,
      "eval_steps_per_second": 0.704,
      "step": 1968
    },
    {
      "epoch": 12.042682926829269,
      "grad_norm": 1.631927728652954,
      "learning_rate": 2.210365853658537e-05,
      "loss": 0.2872,
      "step": 1975
    },
    {
      "epoch": 12.195121951219512,
      "grad_norm": 1.0261545181274414,
      "learning_rate": 2.1680216802168024e-05,
      "loss": 0.2878,
      "step": 2000
    },
    {
      "epoch": 12.347560975609756,
      "grad_norm": 1.1379975080490112,
      "learning_rate": 2.1256775067750677e-05,
      "loss": 0.2945,
      "step": 2025
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.715170681476593,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.2876,
      "step": 2050
    },
    {
      "epoch": 12.652439024390244,
      "grad_norm": 0.5682547092437744,
      "learning_rate": 2.040989159891599e-05,
      "loss": 0.2888,
      "step": 2075
    },
    {
      "epoch": 12.804878048780488,
      "grad_norm": 0.7756302952766418,
      "learning_rate": 1.9986449864498648e-05,
      "loss": 0.2829,
      "step": 2100
    },
    {
      "epoch": 12.957317073170731,
      "grad_norm": 0.8148170709609985,
      "learning_rate": 1.95630081300813e-05,
      "loss": 0.2814,
      "step": 2125
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.8232104121475055,
      "eval_loss": 0.6169049143791199,
      "eval_runtime": 41.1163,
      "eval_samples_per_second": 22.424,
      "eval_steps_per_second": 0.705,
      "step": 2132
    },
    {
      "epoch": 13.109756097560975,
      "grad_norm": 0.6076212525367737,
      "learning_rate": 1.9139566395663956e-05,
      "loss": 0.2877,
      "step": 2150
    },
    {
      "epoch": 13.262195121951219,
      "grad_norm": 0.384480357170105,
      "learning_rate": 1.8716124661246615e-05,
      "loss": 0.2902,
      "step": 2175
    },
    {
      "epoch": 13.414634146341463,
      "grad_norm": 0.4118620455265045,
      "learning_rate": 1.8292682926829268e-05,
      "loss": 0.2849,
      "step": 2200
    },
    {
      "epoch": 13.567073170731707,
      "grad_norm": 0.7292855978012085,
      "learning_rate": 1.7869241192411927e-05,
      "loss": 0.2788,
      "step": 2225
    },
    {
      "epoch": 13.71951219512195,
      "grad_norm": 0.9210421442985535,
      "learning_rate": 1.744579945799458e-05,
      "loss": 0.2804,
      "step": 2250
    },
    {
      "epoch": 13.871951219512194,
      "grad_norm": 1.8374487161636353,
      "learning_rate": 1.702235772357724e-05,
      "loss": 0.2879,
      "step": 2275
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.834056399132321,
      "eval_loss": 0.5981398820877075,
      "eval_runtime": 41.4213,
      "eval_samples_per_second": 22.259,
      "eval_steps_per_second": 0.7,
      "step": 2296
    },
    {
      "epoch": 14.024390243902438,
      "grad_norm": 0.3460817039012909,
      "learning_rate": 1.659891598915989e-05,
      "loss": 0.2844,
      "step": 2300
    },
    {
      "epoch": 14.176829268292684,
      "grad_norm": 0.40026307106018066,
      "learning_rate": 1.6175474254742547e-05,
      "loss": 0.2871,
      "step": 2325
    },
    {
      "epoch": 14.329268292682928,
      "grad_norm": 0.6860306859016418,
      "learning_rate": 1.5752032520325206e-05,
      "loss": 0.2865,
      "step": 2350
    },
    {
      "epoch": 14.481707317073171,
      "grad_norm": 0.32384613156318665,
      "learning_rate": 1.532859078590786e-05,
      "loss": 0.2831,
      "step": 2375
    },
    {
      "epoch": 14.634146341463415,
      "grad_norm": 0.9833716750144958,
      "learning_rate": 1.4905149051490516e-05,
      "loss": 0.2801,
      "step": 2400
    },
    {
      "epoch": 14.786585365853659,
      "grad_norm": 0.4402036964893341,
      "learning_rate": 1.448170731707317e-05,
      "loss": 0.2795,
      "step": 2425
    },
    {
      "epoch": 14.939024390243903,
      "grad_norm": 0.3106571435928345,
      "learning_rate": 1.4058265582655828e-05,
      "loss": 0.2865,
      "step": 2450
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.8459869848156182,
      "eval_loss": 0.6005713939666748,
      "eval_runtime": 40.9776,
      "eval_samples_per_second": 22.5,
      "eval_steps_per_second": 0.708,
      "step": 2460
    },
    {
      "epoch": 15.091463414634147,
      "grad_norm": 0.403603732585907,
      "learning_rate": 1.3634823848238485e-05,
      "loss": 0.2725,
      "step": 2475
    },
    {
      "epoch": 15.24390243902439,
      "grad_norm": 0.337605357170105,
      "learning_rate": 1.321138211382114e-05,
      "loss": 0.2903,
      "step": 2500
    },
    {
      "epoch": 15.396341463414634,
      "grad_norm": 0.4802904427051544,
      "learning_rate": 1.2787940379403795e-05,
      "loss": 0.2829,
      "step": 2525
    },
    {
      "epoch": 15.548780487804878,
      "grad_norm": 0.47631722688674927,
      "learning_rate": 1.2364498644986451e-05,
      "loss": 0.2812,
      "step": 2550
    },
    {
      "epoch": 15.701219512195122,
      "grad_norm": 0.28391534090042114,
      "learning_rate": 1.1941056910569107e-05,
      "loss": 0.2779,
      "step": 2575
    },
    {
      "epoch": 15.853658536585366,
      "grad_norm": 0.5211144685745239,
      "learning_rate": 1.1517615176151763e-05,
      "loss": 0.2818,
      "step": 2600
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.8481561822125814,
      "eval_loss": 0.5918110609054565,
      "eval_runtime": 42.2163,
      "eval_samples_per_second": 21.84,
      "eval_steps_per_second": 0.687,
      "step": 2624
    },
    {
      "epoch": 16.00609756097561,
      "grad_norm": 0.24606825411319733,
      "learning_rate": 1.1094173441734418e-05,
      "loss": 0.286,
      "step": 2625
    },
    {
      "epoch": 16.158536585365855,
      "grad_norm": 0.4235256314277649,
      "learning_rate": 1.0670731707317074e-05,
      "loss": 0.2805,
      "step": 2650
    },
    {
      "epoch": 16.3109756097561,
      "grad_norm": 0.6663039922714233,
      "learning_rate": 1.0247289972899728e-05,
      "loss": 0.2795,
      "step": 2675
    },
    {
      "epoch": 16.463414634146343,
      "grad_norm": 0.5833418965339661,
      "learning_rate": 9.823848238482384e-06,
      "loss": 0.2769,
      "step": 2700
    },
    {
      "epoch": 16.615853658536587,
      "grad_norm": 0.7182282209396362,
      "learning_rate": 9.400406504065042e-06,
      "loss": 0.2814,
      "step": 2725
    },
    {
      "epoch": 16.76829268292683,
      "grad_norm": 0.7894881367683411,
      "learning_rate": 8.976964769647697e-06,
      "loss": 0.2759,
      "step": 2750
    },
    {
      "epoch": 16.920731707317074,
      "grad_norm": 0.47403252124786377,
      "learning_rate": 8.553523035230353e-06,
      "loss": 0.2871,
      "step": 2775
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.8449023861171366,
      "eval_loss": 0.598892331123352,
      "eval_runtime": 40.4981,
      "eval_samples_per_second": 22.766,
      "eval_steps_per_second": 0.716,
      "step": 2788
    },
    {
      "epoch": 17.073170731707318,
      "grad_norm": 0.5262817144393921,
      "learning_rate": 8.130081300813009e-06,
      "loss": 0.283,
      "step": 2800
    },
    {
      "epoch": 17.225609756097562,
      "grad_norm": 0.46620771288871765,
      "learning_rate": 7.706639566395663e-06,
      "loss": 0.2782,
      "step": 2825
    },
    {
      "epoch": 17.378048780487806,
      "grad_norm": 0.6587527394294739,
      "learning_rate": 7.283197831978319e-06,
      "loss": 0.2867,
      "step": 2850
    },
    {
      "epoch": 17.53048780487805,
      "grad_norm": 0.4711247384548187,
      "learning_rate": 6.859756097560977e-06,
      "loss": 0.2858,
      "step": 2875
    },
    {
      "epoch": 17.682926829268293,
      "grad_norm": 0.608683705329895,
      "learning_rate": 6.4363143631436324e-06,
      "loss": 0.2795,
      "step": 2900
    },
    {
      "epoch": 17.835365853658537,
      "grad_norm": 0.3389502763748169,
      "learning_rate": 6.0128726287262874e-06,
      "loss": 0.2731,
      "step": 2925
    },
    {
      "epoch": 17.98780487804878,
      "grad_norm": 0.667849600315094,
      "learning_rate": 5.589430894308943e-06,
      "loss": 0.2757,
      "step": 2950
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.8503253796095445,
      "eval_loss": 0.5923976898193359,
      "eval_runtime": 40.6956,
      "eval_samples_per_second": 22.656,
      "eval_steps_per_second": 0.713,
      "step": 2952
    },
    {
      "epoch": 18.140243902439025,
      "grad_norm": 0.37025541067123413,
      "learning_rate": 5.165989159891599e-06,
      "loss": 0.2799,
      "step": 2975
    },
    {
      "epoch": 18.29268292682927,
      "grad_norm": 0.8489726781845093,
      "learning_rate": 4.742547425474255e-06,
      "loss": 0.2794,
      "step": 3000
    },
    {
      "epoch": 18.445121951219512,
      "grad_norm": 0.32317420840263367,
      "learning_rate": 4.319105691056911e-06,
      "loss": 0.277,
      "step": 3025
    },
    {
      "epoch": 18.597560975609756,
      "grad_norm": 0.3602098822593689,
      "learning_rate": 3.8956639566395666e-06,
      "loss": 0.2805,
      "step": 3050
    },
    {
      "epoch": 18.75,
      "grad_norm": 0.43947067856788635,
      "learning_rate": 3.4722222222222224e-06,
      "loss": 0.2761,
      "step": 3075
    },
    {
      "epoch": 18.902439024390244,
      "grad_norm": 0.3334760069847107,
      "learning_rate": 3.0487804878048782e-06,
      "loss": 0.282,
      "step": 3100
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.8459869848156182,
      "eval_loss": 0.5949397087097168,
      "eval_runtime": 41.1431,
      "eval_samples_per_second": 22.41,
      "eval_steps_per_second": 0.705,
      "step": 3116
    },
    {
      "epoch": 19.054878048780488,
      "grad_norm": 0.3394343852996826,
      "learning_rate": 2.625338753387534e-06,
      "loss": 0.2795,
      "step": 3125
    },
    {
      "epoch": 19.20731707317073,
      "grad_norm": 0.40099793672561646,
      "learning_rate": 2.2018970189701894e-06,
      "loss": 0.2868,
      "step": 3150
    },
    {
      "epoch": 19.359756097560975,
      "grad_norm": 0.28367841243743896,
      "learning_rate": 1.7784552845528455e-06,
      "loss": 0.2813,
      "step": 3175
    },
    {
      "epoch": 19.51219512195122,
      "grad_norm": 0.3525221347808838,
      "learning_rate": 1.3550135501355015e-06,
      "loss": 0.2791,
      "step": 3200
    },
    {
      "epoch": 19.664634146341463,
      "grad_norm": 0.5333865284919739,
      "learning_rate": 9.315718157181572e-07,
      "loss": 0.2793,
      "step": 3225
    },
    {
      "epoch": 19.817073170731707,
      "grad_norm": 0.23533563315868378,
      "learning_rate": 5.081300813008131e-07,
      "loss": 0.2744,
      "step": 3250
    },
    {
      "epoch": 19.96951219512195,
      "grad_norm": 0.23040339350700378,
      "learning_rate": 8.468834688346885e-08,
      "loss": 0.2753,
      "step": 3275
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8492407809110629,
      "eval_loss": 0.5943945050239563,
      "eval_runtime": 42.054,
      "eval_samples_per_second": 21.924,
      "eval_steps_per_second": 0.69,
      "step": 3280
    },
    {
      "epoch": 20.0,
      "step": 3280,
      "total_flos": 0.0,
      "train_loss": 0.3706783848564799,
      "train_runtime": 14788.6671,
      "train_samples_per_second": 7.065,
      "train_steps_per_second": 0.222
    }
  ],
  "logging_steps": 25,
  "max_steps": 3280,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "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": 0.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}