{
  "best_metric": 0.8308770060539246,
  "best_model_checkpoint": "./kd_results/apple/mobilevit-x-small_alpha0.5_temp3.0/checkpoint-1710",
  "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": 3.000155448913574,
      "learning_rate": 6.944444444444445e-06,
      "loss": 1.2674,
      "step": 25
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 1.7359473705291748,
      "learning_rate": 1.388888888888889e-05,
      "loss": 1.2146,
      "step": 50
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 2.8404972553253174,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 1.2144,
      "step": 75
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3102766798418972,
      "eval_loss": 1.3722230195999146,
      "eval_runtime": 30.0366,
      "eval_samples_per_second": 33.692,
      "eval_steps_per_second": 0.533,
      "step": 90
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 2.921933889389038,
      "learning_rate": 2.777777777777778e-05,
      "loss": 1.1914,
      "step": 100
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 2.544729232788086,
      "learning_rate": 3.472222222222222e-05,
      "loss": 1.1965,
      "step": 125
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 1.603236198425293,
      "learning_rate": 4.166666666666667e-05,
      "loss": 1.2031,
      "step": 150
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 3.675070285797119,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 1.1701,
      "step": 175
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3932806324110672,
      "eval_loss": 1.3058674335479736,
      "eval_runtime": 28.6755,
      "eval_samples_per_second": 35.291,
      "eval_steps_per_second": 0.558,
      "step": 180
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 4.072918891906738,
      "learning_rate": 4.938271604938271e-05,
      "loss": 1.1221,
      "step": 200
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.289381265640259,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 1.1079,
      "step": 225
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 2.719942092895508,
      "learning_rate": 4.783950617283951e-05,
      "loss": 1.0635,
      "step": 250
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4792490118577075,
      "eval_loss": 1.1637977361679077,
      "eval_runtime": 28.8919,
      "eval_samples_per_second": 35.027,
      "eval_steps_per_second": 0.554,
      "step": 270
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 3.179309368133545,
      "learning_rate": 4.70679012345679e-05,
      "loss": 1.0444,
      "step": 275
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 2.9999237060546875,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.961,
      "step": 300
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 4.1463189125061035,
      "learning_rate": 4.5524691358024696e-05,
      "loss": 0.9466,
      "step": 325
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 3.4898221492767334,
      "learning_rate": 4.4753086419753084e-05,
      "loss": 0.9234,
      "step": 350
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5217391304347826,
      "eval_loss": 1.0531723499298096,
      "eval_runtime": 28.8945,
      "eval_samples_per_second": 35.024,
      "eval_steps_per_second": 0.554,
      "step": 360
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 4.255654335021973,
      "learning_rate": 4.3981481481481486e-05,
      "loss": 0.8821,
      "step": 375
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 4.979427814483643,
      "learning_rate": 4.3209876543209875e-05,
      "loss": 0.8265,
      "step": 400
    },
    {
      "epoch": 4.722222222222222,
      "grad_norm": 4.477235317230225,
      "learning_rate": 4.243827160493827e-05,
      "loss": 0.8135,
      "step": 425
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.568112850189209,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.8136,
      "step": 450
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5711462450592886,
      "eval_loss": 0.9653530120849609,
      "eval_runtime": 28.5181,
      "eval_samples_per_second": 35.486,
      "eval_steps_per_second": 0.561,
      "step": 450
    },
    {
      "epoch": 5.277777777777778,
      "grad_norm": 3.2484347820281982,
      "learning_rate": 4.089506172839506e-05,
      "loss": 0.7769,
      "step": 475
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 4.245947360992432,
      "learning_rate": 4.012345679012346e-05,
      "loss": 0.7759,
      "step": 500
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 3.700113534927368,
      "learning_rate": 3.935185185185186e-05,
      "loss": 0.7593,
      "step": 525
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.616600790513834,
      "eval_loss": 0.8965586423873901,
      "eval_runtime": 28.7099,
      "eval_samples_per_second": 35.249,
      "eval_steps_per_second": 0.557,
      "step": 540
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 4.6994781494140625,
      "learning_rate": 3.8580246913580246e-05,
      "loss": 0.7306,
      "step": 550
    },
    {
      "epoch": 6.388888888888889,
      "grad_norm": 3.338787794113159,
      "learning_rate": 3.780864197530865e-05,
      "loss": 0.7256,
      "step": 575
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 3.8281936645507812,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.7246,
      "step": 600
    },
    {
      "epoch": 6.944444444444445,
      "grad_norm": 4.3425068855285645,
      "learning_rate": 3.626543209876543e-05,
      "loss": 0.7131,
      "step": 625
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.633399209486166,
      "eval_loss": 0.8710029721260071,
      "eval_runtime": 28.8688,
      "eval_samples_per_second": 35.055,
      "eval_steps_per_second": 0.554,
      "step": 630
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 4.479250907897949,
      "learning_rate": 3.5493827160493834e-05,
      "loss": 0.6762,
      "step": 650
    },
    {
      "epoch": 7.5,
      "grad_norm": 3.777909517288208,
      "learning_rate": 3.472222222222222e-05,
      "loss": 0.6785,
      "step": 675
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 4.595381259918213,
      "learning_rate": 3.395061728395062e-05,
      "loss": 0.6935,
      "step": 700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.599802371541502,
      "eval_loss": 0.931652307510376,
      "eval_runtime": 28.5716,
      "eval_samples_per_second": 35.42,
      "eval_steps_per_second": 0.56,
      "step": 720
    },
    {
      "epoch": 8.055555555555555,
      "grad_norm": 3.7255308628082275,
      "learning_rate": 3.317901234567901e-05,
      "loss": 0.6841,
      "step": 725
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 5.77949333190918,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.6552,
      "step": 750
    },
    {
      "epoch": 8.61111111111111,
      "grad_norm": 3.7085745334625244,
      "learning_rate": 3.16358024691358e-05,
      "loss": 0.6526,
      "step": 775
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 4.154140949249268,
      "learning_rate": 3.08641975308642e-05,
      "loss": 0.6504,
      "step": 800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6294466403162056,
      "eval_loss": 0.8674625158309937,
      "eval_runtime": 28.1732,
      "eval_samples_per_second": 35.921,
      "eval_steps_per_second": 0.568,
      "step": 810
    },
    {
      "epoch": 9.166666666666666,
      "grad_norm": 4.441272735595703,
      "learning_rate": 3.0092592592592593e-05,
      "loss": 0.6353,
      "step": 825
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 6.975021839141846,
      "learning_rate": 2.9320987654320992e-05,
      "loss": 0.6594,
      "step": 850
    },
    {
      "epoch": 9.722222222222221,
      "grad_norm": 7.929254531860352,
      "learning_rate": 2.8549382716049384e-05,
      "loss": 0.6255,
      "step": 875
    },
    {
      "epoch": 10.0,
      "grad_norm": 9.922135353088379,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.6329,
      "step": 900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.6492094861660079,
      "eval_loss": 0.8387980461120605,
      "eval_runtime": 27.9643,
      "eval_samples_per_second": 36.189,
      "eval_steps_per_second": 0.572,
      "step": 900
    },
    {
      "epoch": 10.277777777777779,
      "grad_norm": 4.499290943145752,
      "learning_rate": 2.700617283950617e-05,
      "loss": 0.6044,
      "step": 925
    },
    {
      "epoch": 10.555555555555555,
      "grad_norm": 7.26702356338501,
      "learning_rate": 2.623456790123457e-05,
      "loss": 0.606,
      "step": 950
    },
    {
      "epoch": 10.833333333333334,
      "grad_norm": 4.6892194747924805,
      "learning_rate": 2.5462962962962965e-05,
      "loss": 0.6021,
      "step": 975
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.6521739130434783,
      "eval_loss": 0.8388803601264954,
      "eval_runtime": 28.1159,
      "eval_samples_per_second": 35.994,
      "eval_steps_per_second": 0.569,
      "step": 990
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 5.12283992767334,
      "learning_rate": 2.4691358024691357e-05,
      "loss": 0.6095,
      "step": 1000
    },
    {
      "epoch": 11.38888888888889,
      "grad_norm": 4.385684013366699,
      "learning_rate": 2.3919753086419755e-05,
      "loss": 0.5818,
      "step": 1025
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 8.13125991821289,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.602,
      "step": 1050
    },
    {
      "epoch": 11.944444444444445,
      "grad_norm": 3.8274331092834473,
      "learning_rate": 2.2376543209876542e-05,
      "loss": 0.5937,
      "step": 1075
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.6462450592885376,
      "eval_loss": 0.8454232811927795,
      "eval_runtime": 28.7282,
      "eval_samples_per_second": 35.227,
      "eval_steps_per_second": 0.557,
      "step": 1080
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 4.340463161468506,
      "learning_rate": 2.1604938271604937e-05,
      "loss": 0.5648,
      "step": 1100
    },
    {
      "epoch": 12.5,
      "grad_norm": 3.7954626083374023,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.5718,
      "step": 1125
    },
    {
      "epoch": 12.777777777777779,
      "grad_norm": 4.139441013336182,
      "learning_rate": 2.006172839506173e-05,
      "loss": 0.5687,
      "step": 1150
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.6393280632411067,
      "eval_loss": 0.877737283706665,
      "eval_runtime": 28.0264,
      "eval_samples_per_second": 36.109,
      "eval_steps_per_second": 0.571,
      "step": 1170
    },
    {
      "epoch": 13.055555555555555,
      "grad_norm": 4.385037422180176,
      "learning_rate": 1.9290123456790123e-05,
      "loss": 0.5711,
      "step": 1175
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 4.172731876373291,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.5773,
      "step": 1200
    },
    {
      "epoch": 13.61111111111111,
      "grad_norm": 7.134023189544678,
      "learning_rate": 1.7746913580246917e-05,
      "loss": 0.5533,
      "step": 1225
    },
    {
      "epoch": 13.88888888888889,
      "grad_norm": 5.6738457679748535,
      "learning_rate": 1.697530864197531e-05,
      "loss": 0.5647,
      "step": 1250
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.6353754940711462,
      "eval_loss": 0.8433853387832642,
      "eval_runtime": 28.5988,
      "eval_samples_per_second": 35.386,
      "eval_steps_per_second": 0.559,
      "step": 1260
    },
    {
      "epoch": 14.166666666666666,
      "grad_norm": 6.407425880432129,
      "learning_rate": 1.6203703703703704e-05,
      "loss": 0.5627,
      "step": 1275
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 3.377164125442505,
      "learning_rate": 1.54320987654321e-05,
      "loss": 0.5392,
      "step": 1300
    },
    {
      "epoch": 14.722222222222221,
      "grad_norm": 5.9406914710998535,
      "learning_rate": 1.4660493827160496e-05,
      "loss": 0.5498,
      "step": 1325
    },
    {
      "epoch": 15.0,
      "grad_norm": 9.535811424255371,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.5372,
      "step": 1350
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.6561264822134387,
      "eval_loss": 0.8427901864051819,
      "eval_runtime": 28.6719,
      "eval_samples_per_second": 35.296,
      "eval_steps_per_second": 0.558,
      "step": 1350
    },
    {
      "epoch": 15.277777777777779,
      "grad_norm": 4.922372341156006,
      "learning_rate": 1.3117283950617285e-05,
      "loss": 0.5333,
      "step": 1375
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 3.7180538177490234,
      "learning_rate": 1.2345679012345678e-05,
      "loss": 0.5403,
      "step": 1400
    },
    {
      "epoch": 15.833333333333334,
      "grad_norm": 5.1782732009887695,
      "learning_rate": 1.1574074074074075e-05,
      "loss": 0.5302,
      "step": 1425
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.6511857707509882,
      "eval_loss": 0.8375866413116455,
      "eval_runtime": 28.4207,
      "eval_samples_per_second": 35.608,
      "eval_steps_per_second": 0.563,
      "step": 1440
    },
    {
      "epoch": 16.11111111111111,
      "grad_norm": 3.6969380378723145,
      "learning_rate": 1.0802469135802469e-05,
      "loss": 0.5183,
      "step": 1450
    },
    {
      "epoch": 16.38888888888889,
      "grad_norm": 5.547866344451904,
      "learning_rate": 1.0030864197530866e-05,
      "loss": 0.522,
      "step": 1475
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 4.255801200866699,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.5335,
      "step": 1500
    },
    {
      "epoch": 16.944444444444443,
      "grad_norm": 5.837561130523682,
      "learning_rate": 8.487654320987654e-06,
      "loss": 0.5374,
      "step": 1525
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.6561264822134387,
      "eval_loss": 0.8364587426185608,
      "eval_runtime": 28.7115,
      "eval_samples_per_second": 35.247,
      "eval_steps_per_second": 0.557,
      "step": 1530
    },
    {
      "epoch": 17.22222222222222,
      "grad_norm": 5.2990522384643555,
      "learning_rate": 7.71604938271605e-06,
      "loss": 0.5067,
      "step": 1550
    },
    {
      "epoch": 17.5,
      "grad_norm": 4.39614725112915,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.5192,
      "step": 1575
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 3.598612070083618,
      "learning_rate": 6.172839506172839e-06,
      "loss": 0.5273,
      "step": 1600
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.6610671936758893,
      "eval_loss": 0.843079686164856,
      "eval_runtime": 28.4716,
      "eval_samples_per_second": 35.544,
      "eval_steps_per_second": 0.562,
      "step": 1620
    },
    {
      "epoch": 18.055555555555557,
      "grad_norm": 4.549961090087891,
      "learning_rate": 5.401234567901234e-06,
      "loss": 0.5151,
      "step": 1625
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 4.037781238555908,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.5283,
      "step": 1650
    },
    {
      "epoch": 18.61111111111111,
      "grad_norm": 6.193184852600098,
      "learning_rate": 3.858024691358025e-06,
      "loss": 0.5122,
      "step": 1675
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 4.468006610870361,
      "learning_rate": 3.0864197530864196e-06,
      "loss": 0.5197,
      "step": 1700
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.658102766798419,
      "eval_loss": 0.8308770060539246,
      "eval_runtime": 28.9316,
      "eval_samples_per_second": 34.979,
      "eval_steps_per_second": 0.553,
      "step": 1710
    },
    {
      "epoch": 19.166666666666668,
      "grad_norm": 3.8190793991088867,
      "learning_rate": 2.3148148148148148e-06,
      "loss": 0.5149,
      "step": 1725
    },
    {
      "epoch": 19.444444444444443,
      "grad_norm": 4.120711326599121,
      "learning_rate": 1.5432098765432098e-06,
      "loss": 0.5045,
      "step": 1750
    },
    {
      "epoch": 19.72222222222222,
      "grad_norm": 7.288771152496338,
      "learning_rate": 7.716049382716049e-07,
      "loss": 0.5266,
      "step": 1775
    },
    {
      "epoch": 20.0,
      "grad_norm": 9.75727653503418,
      "learning_rate": 0.0,
      "loss": 0.5123,
      "step": 1800
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.6669960474308301,
      "eval_loss": 0.8315740823745728,
      "eval_runtime": 28.4521,
      "eval_samples_per_second": 35.569,
      "eval_steps_per_second": 0.562,
      "step": 1800
    },
    {
      "epoch": 20.0,
      "step": 1800,
      "total_flos": 0.0,
      "train_loss": 0.7107453621758355,
      "train_runtime": 10382.202,
      "train_samples_per_second": 11.038,
      "train_steps_per_second": 0.173
    }
  ],
  "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
}