{
  "best_metric": 0.5586743354797363,
  "best_model_checkpoint": "./kd_results/microsoft/beit-base-patch16-224_alpha0.7_temp5.0/checkpoint-1440",
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 1800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 4.854953765869141,
      "learning_rate": 6.944444444444445e-06,
      "loss": 1.1971,
      "step": 25
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 4.132758140563965,
      "learning_rate": 1.388888888888889e-05,
      "loss": 1.1345,
      "step": 50
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 7.529750347137451,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 1.1179,
      "step": 75
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.2875494071146245,
      "eval_loss": 1.4253844022750854,
      "eval_runtime": 31.7768,
      "eval_samples_per_second": 31.847,
      "eval_steps_per_second": 0.504,
      "step": 90
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 5.156152248382568,
      "learning_rate": 2.777777777777778e-05,
      "loss": 1.0713,
      "step": 100
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 3.974464178085327,
      "learning_rate": 3.472222222222222e-05,
      "loss": 1.0127,
      "step": 125
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 6.480051517486572,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.9649,
      "step": 150
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 5.899675369262695,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 0.8879,
      "step": 175
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5632411067193676,
      "eval_loss": 1.0517464876174927,
      "eval_runtime": 31.398,
      "eval_samples_per_second": 32.231,
      "eval_steps_per_second": 0.51,
      "step": 180
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 8.341927528381348,
      "learning_rate": 4.938271604938271e-05,
      "loss": 0.6881,
      "step": 200
    },
    {
      "epoch": 2.5,
      "grad_norm": 5.7711968421936035,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 0.6336,
      "step": 225
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 5.359498500823975,
      "learning_rate": 4.783950617283951e-05,
      "loss": 0.5644,
      "step": 250
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7430830039525692,
      "eval_loss": 0.7063300609588623,
      "eval_runtime": 31.3401,
      "eval_samples_per_second": 32.291,
      "eval_steps_per_second": 0.511,
      "step": 270
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 3.9833033084869385,
      "learning_rate": 4.70679012345679e-05,
      "loss": 0.4968,
      "step": 275
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 4.732687473297119,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.4054,
      "step": 300
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 6.728626251220703,
      "learning_rate": 4.5524691358024696e-05,
      "loss": 0.3964,
      "step": 325
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 4.665893077850342,
      "learning_rate": 4.4753086419753084e-05,
      "loss": 0.3909,
      "step": 350
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 0.6420592069625854,
      "eval_runtime": 30.3488,
      "eval_samples_per_second": 33.346,
      "eval_steps_per_second": 0.527,
      "step": 360
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 4.054627418518066,
      "learning_rate": 4.3981481481481486e-05,
      "loss": 0.3229,
      "step": 375
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 4.525479316711426,
      "learning_rate": 4.3209876543209875e-05,
      "loss": 0.2746,
      "step": 400
    },
    {
      "epoch": 4.722222222222222,
      "grad_norm": 3.628028392791748,
      "learning_rate": 4.243827160493827e-05,
      "loss": 0.2708,
      "step": 425
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.1524553298950195,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.2787,
      "step": 450
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7875494071146245,
      "eval_loss": 0.6018449664115906,
      "eval_runtime": 30.9616,
      "eval_samples_per_second": 32.686,
      "eval_steps_per_second": 0.517,
      "step": 450
    },
    {
      "epoch": 5.277777777777778,
      "grad_norm": 3.271401882171631,
      "learning_rate": 4.089506172839506e-05,
      "loss": 0.2339,
      "step": 475
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 3.5659079551696777,
      "learning_rate": 4.012345679012346e-05,
      "loss": 0.2227,
      "step": 500
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 3.4077155590057373,
      "learning_rate": 3.935185185185186e-05,
      "loss": 0.2167,
      "step": 525
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7687747035573123,
      "eval_loss": 0.657771646976471,
      "eval_runtime": 31.3505,
      "eval_samples_per_second": 32.28,
      "eval_steps_per_second": 0.51,
      "step": 540
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 2.961601734161377,
      "learning_rate": 3.8580246913580246e-05,
      "loss": 0.2173,
      "step": 550
    },
    {
      "epoch": 6.388888888888889,
      "grad_norm": 3.4833714962005615,
      "learning_rate": 3.780864197530865e-05,
      "loss": 0.2049,
      "step": 575
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 3.0653741359710693,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.2048,
      "step": 600
    },
    {
      "epoch": 6.944444444444445,
      "grad_norm": 2.5728049278259277,
      "learning_rate": 3.626543209876543e-05,
      "loss": 0.1982,
      "step": 625
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.782608695652174,
      "eval_loss": 0.6106678247451782,
      "eval_runtime": 32.139,
      "eval_samples_per_second": 31.488,
      "eval_steps_per_second": 0.498,
      "step": 630
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 3.3970484733581543,
      "learning_rate": 3.5493827160493834e-05,
      "loss": 0.1764,
      "step": 650
    },
    {
      "epoch": 7.5,
      "grad_norm": 2.866818904876709,
      "learning_rate": 3.472222222222222e-05,
      "loss": 0.1784,
      "step": 675
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 2.679239273071289,
      "learning_rate": 3.395061728395062e-05,
      "loss": 0.1799,
      "step": 700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8132411067193676,
      "eval_loss": 0.5741682648658752,
      "eval_runtime": 31.3562,
      "eval_samples_per_second": 32.274,
      "eval_steps_per_second": 0.51,
      "step": 720
    },
    {
      "epoch": 8.055555555555555,
      "grad_norm": 2.2661197185516357,
      "learning_rate": 3.317901234567901e-05,
      "loss": 0.1873,
      "step": 725
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 1.946643590927124,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.1704,
      "step": 750
    },
    {
      "epoch": 8.61111111111111,
      "grad_norm": 1.6869217157363892,
      "learning_rate": 3.16358024691358e-05,
      "loss": 0.1652,
      "step": 775
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 3.4906277656555176,
      "learning_rate": 3.08641975308642e-05,
      "loss": 0.1689,
      "step": 800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.7944664031620553,
      "eval_loss": 0.6021401882171631,
      "eval_runtime": 31.727,
      "eval_samples_per_second": 31.897,
      "eval_steps_per_second": 0.504,
      "step": 810
    },
    {
      "epoch": 9.166666666666666,
      "grad_norm": 2.9145889282226562,
      "learning_rate": 3.0092592592592593e-05,
      "loss": 0.1589,
      "step": 825
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 1.5670665502548218,
      "learning_rate": 2.9320987654320992e-05,
      "loss": 0.1537,
      "step": 850
    },
    {
      "epoch": 9.722222222222221,
      "grad_norm": 1.8959583044052124,
      "learning_rate": 2.8549382716049384e-05,
      "loss": 0.1528,
      "step": 875
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.9776724576950073,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.1555,
      "step": 900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.7895256916996047,
      "eval_loss": 0.587110698223114,
      "eval_runtime": 30.4121,
      "eval_samples_per_second": 33.276,
      "eval_steps_per_second": 0.526,
      "step": 900
    },
    {
      "epoch": 10.277777777777779,
      "grad_norm": 2.025183916091919,
      "learning_rate": 2.700617283950617e-05,
      "loss": 0.1562,
      "step": 925
    },
    {
      "epoch": 10.555555555555555,
      "grad_norm": 1.173431634902954,
      "learning_rate": 2.623456790123457e-05,
      "loss": 0.1538,
      "step": 950
    },
    {
      "epoch": 10.833333333333334,
      "grad_norm": 3.4991841316223145,
      "learning_rate": 2.5462962962962965e-05,
      "loss": 0.1514,
      "step": 975
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.791501976284585,
      "eval_loss": 0.5836030840873718,
      "eval_runtime": 32.0032,
      "eval_samples_per_second": 31.622,
      "eval_steps_per_second": 0.5,
      "step": 990
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 1.008795142173767,
      "learning_rate": 2.4691358024691357e-05,
      "loss": 0.1416,
      "step": 1000
    },
    {
      "epoch": 11.38888888888889,
      "grad_norm": 3.103557586669922,
      "learning_rate": 2.3919753086419755e-05,
      "loss": 0.1361,
      "step": 1025
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 1.1227365732192993,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.1444,
      "step": 1050
    },
    {
      "epoch": 11.944444444444445,
      "grad_norm": 1.0779048204421997,
      "learning_rate": 2.2376543209876542e-05,
      "loss": 0.1418,
      "step": 1075
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.7964426877470355,
      "eval_loss": 0.5684204697608948,
      "eval_runtime": 31.8178,
      "eval_samples_per_second": 31.806,
      "eval_steps_per_second": 0.503,
      "step": 1080
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 1.3879923820495605,
      "learning_rate": 2.1604938271604937e-05,
      "loss": 0.1366,
      "step": 1100
    },
    {
      "epoch": 12.5,
      "grad_norm": 2.631446123123169,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.145,
      "step": 1125
    },
    {
      "epoch": 12.777777777777779,
      "grad_norm": 1.33954656124115,
      "learning_rate": 2.006172839506173e-05,
      "loss": 0.1356,
      "step": 1150
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.7974308300395256,
      "eval_loss": 0.568759024143219,
      "eval_runtime": 31.3892,
      "eval_samples_per_second": 32.24,
      "eval_steps_per_second": 0.51,
      "step": 1170
    },
    {
      "epoch": 13.055555555555555,
      "grad_norm": 1.1815236806869507,
      "learning_rate": 1.9290123456790123e-05,
      "loss": 0.1306,
      "step": 1175
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 1.3230443000793457,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.1382,
      "step": 1200
    },
    {
      "epoch": 13.61111111111111,
      "grad_norm": 1.0125420093536377,
      "learning_rate": 1.7746913580246917e-05,
      "loss": 0.1309,
      "step": 1225
    },
    {
      "epoch": 13.88888888888889,
      "grad_norm": 0.9248207807540894,
      "learning_rate": 1.697530864197531e-05,
      "loss": 0.1262,
      "step": 1250
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8053359683794467,
      "eval_loss": 0.5782861709594727,
      "eval_runtime": 31.6992,
      "eval_samples_per_second": 31.925,
      "eval_steps_per_second": 0.505,
      "step": 1260
    },
    {
      "epoch": 14.166666666666666,
      "grad_norm": 1.0868349075317383,
      "learning_rate": 1.6203703703703704e-05,
      "loss": 0.1308,
      "step": 1275
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 0.8695051670074463,
      "learning_rate": 1.54320987654321e-05,
      "loss": 0.129,
      "step": 1300
    },
    {
      "epoch": 14.722222222222221,
      "grad_norm": 1.301735520362854,
      "learning_rate": 1.4660493827160496e-05,
      "loss": 0.1255,
      "step": 1325
    },
    {
      "epoch": 15.0,
      "grad_norm": 2.268371105194092,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.128,
      "step": 1350
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.7924901185770751,
      "eval_loss": 0.5665250420570374,
      "eval_runtime": 31.6739,
      "eval_samples_per_second": 31.951,
      "eval_steps_per_second": 0.505,
      "step": 1350
    },
    {
      "epoch": 15.277777777777779,
      "grad_norm": 0.8186533451080322,
      "learning_rate": 1.3117283950617285e-05,
      "loss": 0.1268,
      "step": 1375
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 1.092484474182129,
      "learning_rate": 1.2345679012345678e-05,
      "loss": 0.1225,
      "step": 1400
    },
    {
      "epoch": 15.833333333333334,
      "grad_norm": 0.8935237526893616,
      "learning_rate": 1.1574074074074075e-05,
      "loss": 0.1211,
      "step": 1425
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.8013833992094862,
      "eval_loss": 0.5586743354797363,
      "eval_runtime": 31.5109,
      "eval_samples_per_second": 32.116,
      "eval_steps_per_second": 0.508,
      "step": 1440
    },
    {
      "epoch": 16.11111111111111,
      "grad_norm": 0.5855790376663208,
      "learning_rate": 1.0802469135802469e-05,
      "loss": 0.1195,
      "step": 1450
    },
    {
      "epoch": 16.38888888888889,
      "grad_norm": 0.7036534547805786,
      "learning_rate": 1.0030864197530866e-05,
      "loss": 0.1181,
      "step": 1475
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 0.7655908465385437,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.1217,
      "step": 1500
    },
    {
      "epoch": 16.944444444444443,
      "grad_norm": 1.5623362064361572,
      "learning_rate": 8.487654320987654e-06,
      "loss": 0.1223,
      "step": 1525
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.8033596837944664,
      "eval_loss": 0.5643883347511292,
      "eval_runtime": 31.1811,
      "eval_samples_per_second": 32.456,
      "eval_steps_per_second": 0.513,
      "step": 1530
    },
    {
      "epoch": 17.22222222222222,
      "grad_norm": 0.7560245394706726,
      "learning_rate": 7.71604938271605e-06,
      "loss": 0.1207,
      "step": 1550
    },
    {
      "epoch": 17.5,
      "grad_norm": 0.9252073764801025,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.1168,
      "step": 1575
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 0.8090717196464539,
      "learning_rate": 6.172839506172839e-06,
      "loss": 0.1133,
      "step": 1600
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.808300395256917,
      "eval_loss": 0.5634317398071289,
      "eval_runtime": 31.5051,
      "eval_samples_per_second": 32.122,
      "eval_steps_per_second": 0.508,
      "step": 1620
    },
    {
      "epoch": 18.055555555555557,
      "grad_norm": 0.8598595261573792,
      "learning_rate": 5.401234567901234e-06,
      "loss": 0.122,
      "step": 1625
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 0.6354872584342957,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.1203,
      "step": 1650
    },
    {
      "epoch": 18.61111111111111,
      "grad_norm": 0.6701659560203552,
      "learning_rate": 3.858024691358025e-06,
      "loss": 0.1113,
      "step": 1675
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 0.6543203592300415,
      "learning_rate": 3.0864197530864196e-06,
      "loss": 0.1158,
      "step": 1700
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.8063241106719368,
      "eval_loss": 0.5711103677749634,
      "eval_runtime": 31.1447,
      "eval_samples_per_second": 32.493,
      "eval_steps_per_second": 0.514,
      "step": 1710
    },
    {
      "epoch": 19.166666666666668,
      "grad_norm": 0.8226360082626343,
      "learning_rate": 2.3148148148148148e-06,
      "loss": 0.113,
      "step": 1725
    },
    {
      "epoch": 19.444444444444443,
      "grad_norm": 0.611422061920166,
      "learning_rate": 1.5432098765432098e-06,
      "loss": 0.1178,
      "step": 1750
    },
    {
      "epoch": 19.72222222222222,
      "grad_norm": 0.5451157689094543,
      "learning_rate": 7.716049382716049e-07,
      "loss": 0.1157,
      "step": 1775
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.8388997912406921,
      "learning_rate": 0.0,
      "loss": 0.1119,
      "step": 1800
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.808300395256917,
      "eval_loss": 0.5648430585861206,
      "eval_runtime": 32.2673,
      "eval_samples_per_second": 31.363,
      "eval_steps_per_second": 0.496,
      "step": 1800
    },
    {
      "epoch": 20.0,
      "step": 1800,
      "total_flos": 0.0,
      "train_loss": 0.2787140127023061,
      "train_runtime": 10963.1372,
      "train_samples_per_second": 10.453,
      "train_steps_per_second": 0.164
    }
  ],
  "logging_steps": 25,
  "max_steps": 1800,
  "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": 64,
  "trial_name": null,
  "trial_params": null
}