{
  "best_metric": 0.8549718856811523,
  "best_model_checkpoint": "./kd_results/apple/mobilevit-small_alpha0.7_temp5.0/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": 5.067017078399658,
      "learning_rate": 6.944444444444445e-06,
      "loss": 1.3328,
      "step": 25
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 1.9458526372909546,
      "learning_rate": 1.388888888888889e-05,
      "loss": 1.136,
      "step": 50
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 1.8754066228866577,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 1.1311,
      "step": 75
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.24703557312252963,
      "eval_loss": 1.4279208183288574,
      "eval_runtime": 28.5592,
      "eval_samples_per_second": 35.435,
      "eval_steps_per_second": 0.56,
      "step": 90
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 2.692765951156616,
      "learning_rate": 2.777777777777778e-05,
      "loss": 1.1094,
      "step": 100
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 3.097252607345581,
      "learning_rate": 3.472222222222222e-05,
      "loss": 1.1009,
      "step": 125
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 1.8249784708023071,
      "learning_rate": 4.166666666666667e-05,
      "loss": 1.1182,
      "step": 150
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 3.1942126750946045,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 1.0789,
      "step": 175
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3191699604743083,
      "eval_loss": 1.333102822303772,
      "eval_runtime": 28.5466,
      "eval_samples_per_second": 35.451,
      "eval_steps_per_second": 0.56,
      "step": 180
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 5.066985607147217,
      "learning_rate": 4.938271604938271e-05,
      "loss": 1.0393,
      "step": 200
    },
    {
      "epoch": 2.5,
      "grad_norm": 4.512152671813965,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 1.0193,
      "step": 225
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 4.158876895904541,
      "learning_rate": 4.783950617283951e-05,
      "loss": 0.9867,
      "step": 250
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.38636363636363635,
      "eval_loss": 1.2525354623794556,
      "eval_runtime": 28.5695,
      "eval_samples_per_second": 35.422,
      "eval_steps_per_second": 0.56,
      "step": 270
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 4.744362831115723,
      "learning_rate": 4.70679012345679e-05,
      "loss": 0.961,
      "step": 275
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 4.1431050300598145,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.8868,
      "step": 300
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 3.916766405105591,
      "learning_rate": 4.5524691358024696e-05,
      "loss": 0.8619,
      "step": 325
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 3.961512804031372,
      "learning_rate": 4.4753086419753084e-05,
      "loss": 0.8433,
      "step": 350
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.49703557312252966,
      "eval_loss": 1.067132830619812,
      "eval_runtime": 28.8584,
      "eval_samples_per_second": 35.068,
      "eval_steps_per_second": 0.554,
      "step": 360
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 4.347306728363037,
      "learning_rate": 4.3981481481481486e-05,
      "loss": 0.7999,
      "step": 375
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 5.401865005493164,
      "learning_rate": 4.3209876543209875e-05,
      "loss": 0.7392,
      "step": 400
    },
    {
      "epoch": 4.722222222222222,
      "grad_norm": 5.163962364196777,
      "learning_rate": 4.243827160493827e-05,
      "loss": 0.725,
      "step": 425
    },
    {
      "epoch": 5.0,
      "grad_norm": 6.064295291900635,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.7496,
      "step": 450
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5602766798418972,
      "eval_loss": 0.9999355673789978,
      "eval_runtime": 28.6185,
      "eval_samples_per_second": 35.362,
      "eval_steps_per_second": 0.559,
      "step": 450
    },
    {
      "epoch": 5.277777777777778,
      "grad_norm": 3.596616744995117,
      "learning_rate": 4.089506172839506e-05,
      "loss": 0.7087,
      "step": 475
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 8.382354736328125,
      "learning_rate": 4.012345679012346e-05,
      "loss": 0.7061,
      "step": 500
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 5.777190685272217,
      "learning_rate": 3.935185185185186e-05,
      "loss": 0.6948,
      "step": 525
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5988142292490118,
      "eval_loss": 0.937413215637207,
      "eval_runtime": 28.9208,
      "eval_samples_per_second": 34.992,
      "eval_steps_per_second": 0.553,
      "step": 540
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 4.378117561340332,
      "learning_rate": 3.8580246913580246e-05,
      "loss": 0.6473,
      "step": 550
    },
    {
      "epoch": 6.388888888888889,
      "grad_norm": 6.292396545410156,
      "learning_rate": 3.780864197530865e-05,
      "loss": 0.6451,
      "step": 575
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 10.105191230773926,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.6372,
      "step": 600
    },
    {
      "epoch": 6.944444444444445,
      "grad_norm": 4.2663350105285645,
      "learning_rate": 3.626543209876543e-05,
      "loss": 0.6495,
      "step": 625
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6146245059288538,
      "eval_loss": 0.9065309166908264,
      "eval_runtime": 28.1886,
      "eval_samples_per_second": 35.901,
      "eval_steps_per_second": 0.568,
      "step": 630
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 6.072882175445557,
      "learning_rate": 3.5493827160493834e-05,
      "loss": 0.5943,
      "step": 650
    },
    {
      "epoch": 7.5,
      "grad_norm": 7.997155666351318,
      "learning_rate": 3.472222222222222e-05,
      "loss": 0.592,
      "step": 675
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 5.652843952178955,
      "learning_rate": 3.395061728395062e-05,
      "loss": 0.6101,
      "step": 700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.6007905138339921,
      "eval_loss": 0.9330638647079468,
      "eval_runtime": 28.9448,
      "eval_samples_per_second": 34.963,
      "eval_steps_per_second": 0.553,
      "step": 720
    },
    {
      "epoch": 8.055555555555555,
      "grad_norm": 5.346714019775391,
      "learning_rate": 3.317901234567901e-05,
      "loss": 0.5961,
      "step": 725
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 5.7756123542785645,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.5641,
      "step": 750
    },
    {
      "epoch": 8.61111111111111,
      "grad_norm": 4.24644136428833,
      "learning_rate": 3.16358024691358e-05,
      "loss": 0.5661,
      "step": 775
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 7.723337173461914,
      "learning_rate": 3.08641975308642e-05,
      "loss": 0.5591,
      "step": 800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6571146245059288,
      "eval_loss": 0.8673222064971924,
      "eval_runtime": 29.0659,
      "eval_samples_per_second": 34.817,
      "eval_steps_per_second": 0.55,
      "step": 810
    },
    {
      "epoch": 9.166666666666666,
      "grad_norm": 6.067270755767822,
      "learning_rate": 3.0092592592592593e-05,
      "loss": 0.5474,
      "step": 825
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 4.425726413726807,
      "learning_rate": 2.9320987654320992e-05,
      "loss": 0.5453,
      "step": 850
    },
    {
      "epoch": 9.722222222222221,
      "grad_norm": 10.190550804138184,
      "learning_rate": 2.8549382716049384e-05,
      "loss": 0.5452,
      "step": 875
    },
    {
      "epoch": 10.0,
      "grad_norm": 8.709136009216309,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.5406,
      "step": 900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.6235177865612648,
      "eval_loss": 0.8704849481582642,
      "eval_runtime": 28.8308,
      "eval_samples_per_second": 35.101,
      "eval_steps_per_second": 0.555,
      "step": 900
    },
    {
      "epoch": 10.277777777777779,
      "grad_norm": 2.9625649452209473,
      "learning_rate": 2.700617283950617e-05,
      "loss": 0.5191,
      "step": 925
    },
    {
      "epoch": 10.555555555555555,
      "grad_norm": 5.884112358093262,
      "learning_rate": 2.623456790123457e-05,
      "loss": 0.5156,
      "step": 950
    },
    {
      "epoch": 10.833333333333334,
      "grad_norm": 9.00991153717041,
      "learning_rate": 2.5462962962962965e-05,
      "loss": 0.5188,
      "step": 975
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.6630434782608695,
      "eval_loss": 0.8578750491142273,
      "eval_runtime": 28.2136,
      "eval_samples_per_second": 35.869,
      "eval_steps_per_second": 0.567,
      "step": 990
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 6.289473533630371,
      "learning_rate": 2.4691358024691357e-05,
      "loss": 0.5064,
      "step": 1000
    },
    {
      "epoch": 11.38888888888889,
      "grad_norm": 5.463348388671875,
      "learning_rate": 2.3919753086419755e-05,
      "loss": 0.4854,
      "step": 1025
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 5.968869686126709,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.4956,
      "step": 1050
    },
    {
      "epoch": 11.944444444444445,
      "grad_norm": 3.9279537200927734,
      "learning_rate": 2.2376543209876542e-05,
      "loss": 0.4875,
      "step": 1075
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.6324110671936759,
      "eval_loss": 0.8808870315551758,
      "eval_runtime": 28.7749,
      "eval_samples_per_second": 35.17,
      "eval_steps_per_second": 0.556,
      "step": 1080
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 2.7190964221954346,
      "learning_rate": 2.1604938271604937e-05,
      "loss": 0.4716,
      "step": 1100
    },
    {
      "epoch": 12.5,
      "grad_norm": 5.235175132751465,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.4614,
      "step": 1125
    },
    {
      "epoch": 12.777777777777779,
      "grad_norm": 4.206543445587158,
      "learning_rate": 2.006172839506173e-05,
      "loss": 0.4596,
      "step": 1150
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.6324110671936759,
      "eval_loss": 0.8903521299362183,
      "eval_runtime": 28.7353,
      "eval_samples_per_second": 35.218,
      "eval_steps_per_second": 0.557,
      "step": 1170
    },
    {
      "epoch": 13.055555555555555,
      "grad_norm": 5.413076400756836,
      "learning_rate": 1.9290123456790123e-05,
      "loss": 0.4607,
      "step": 1175
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 7.035764217376709,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.4549,
      "step": 1200
    },
    {
      "epoch": 13.61111111111111,
      "grad_norm": 7.080498218536377,
      "learning_rate": 1.7746913580246917e-05,
      "loss": 0.445,
      "step": 1225
    },
    {
      "epoch": 13.88888888888889,
      "grad_norm": 5.1540937423706055,
      "learning_rate": 1.697530864197531e-05,
      "loss": 0.4418,
      "step": 1250
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.6541501976284585,
      "eval_loss": 0.8596166968345642,
      "eval_runtime": 28.5111,
      "eval_samples_per_second": 35.495,
      "eval_steps_per_second": 0.561,
      "step": 1260
    },
    {
      "epoch": 14.166666666666666,
      "grad_norm": 4.73467493057251,
      "learning_rate": 1.6203703703703704e-05,
      "loss": 0.4534,
      "step": 1275
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 3.350318193435669,
      "learning_rate": 1.54320987654321e-05,
      "loss": 0.4293,
      "step": 1300
    },
    {
      "epoch": 14.722222222222221,
      "grad_norm": 5.757190227508545,
      "learning_rate": 1.4660493827160496e-05,
      "loss": 0.437,
      "step": 1325
    },
    {
      "epoch": 15.0,
      "grad_norm": 10.14773178100586,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.4385,
      "step": 1350
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.6304347826086957,
      "eval_loss": 0.8768613338470459,
      "eval_runtime": 28.9356,
      "eval_samples_per_second": 34.974,
      "eval_steps_per_second": 0.553,
      "step": 1350
    },
    {
      "epoch": 15.277777777777779,
      "grad_norm": 4.023919105529785,
      "learning_rate": 1.3117283950617285e-05,
      "loss": 0.418,
      "step": 1375
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 5.213397026062012,
      "learning_rate": 1.2345679012345678e-05,
      "loss": 0.4343,
      "step": 1400
    },
    {
      "epoch": 15.833333333333334,
      "grad_norm": 5.7322797775268555,
      "learning_rate": 1.1574074074074075e-05,
      "loss": 0.4175,
      "step": 1425
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.6561264822134387,
      "eval_loss": 0.8624676465988159,
      "eval_runtime": 28.9613,
      "eval_samples_per_second": 34.943,
      "eval_steps_per_second": 0.552,
      "step": 1440
    },
    {
      "epoch": 16.11111111111111,
      "grad_norm": 2.6554200649261475,
      "learning_rate": 1.0802469135802469e-05,
      "loss": 0.4027,
      "step": 1450
    },
    {
      "epoch": 16.38888888888889,
      "grad_norm": 4.859523773193359,
      "learning_rate": 1.0030864197530866e-05,
      "loss": 0.4188,
      "step": 1475
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 6.244664669036865,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.4194,
      "step": 1500
    },
    {
      "epoch": 16.944444444444443,
      "grad_norm": 4.917417049407959,
      "learning_rate": 8.487654320987654e-06,
      "loss": 0.413,
      "step": 1525
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.6571146245059288,
      "eval_loss": 0.8577585220336914,
      "eval_runtime": 28.9271,
      "eval_samples_per_second": 34.985,
      "eval_steps_per_second": 0.553,
      "step": 1530
    },
    {
      "epoch": 17.22222222222222,
      "grad_norm": 5.908167839050293,
      "learning_rate": 7.71604938271605e-06,
      "loss": 0.4077,
      "step": 1550
    },
    {
      "epoch": 17.5,
      "grad_norm": 4.492807388305664,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.3998,
      "step": 1575
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 3.352818489074707,
      "learning_rate": 6.172839506172839e-06,
      "loss": 0.3973,
      "step": 1600
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.658102766798419,
      "eval_loss": 0.857342541217804,
      "eval_runtime": 29.3279,
      "eval_samples_per_second": 34.506,
      "eval_steps_per_second": 0.546,
      "step": 1620
    },
    {
      "epoch": 18.055555555555557,
      "grad_norm": 3.0056979656219482,
      "learning_rate": 5.401234567901234e-06,
      "loss": 0.405,
      "step": 1625
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 3.8960962295532227,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.4028,
      "step": 1650
    },
    {
      "epoch": 18.61111111111111,
      "grad_norm": 3.9841091632843018,
      "learning_rate": 3.858024691358025e-06,
      "loss": 0.3954,
      "step": 1675
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 3.473726749420166,
      "learning_rate": 3.0864197530864196e-06,
      "loss": 0.3964,
      "step": 1700
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.6679841897233202,
      "eval_loss": 0.8549718856811523,
      "eval_runtime": 28.7243,
      "eval_samples_per_second": 35.232,
      "eval_steps_per_second": 0.557,
      "step": 1710
    },
    {
      "epoch": 19.166666666666668,
      "grad_norm": 3.783536672592163,
      "learning_rate": 2.3148148148148148e-06,
      "loss": 0.401,
      "step": 1725
    },
    {
      "epoch": 19.444444444444443,
      "grad_norm": 2.4962282180786133,
      "learning_rate": 1.5432098765432098e-06,
      "loss": 0.3949,
      "step": 1750
    },
    {
      "epoch": 19.72222222222222,
      "grad_norm": 5.234297752380371,
      "learning_rate": 7.716049382716049e-07,
      "loss": 0.3994,
      "step": 1775
    },
    {
      "epoch": 20.0,
      "grad_norm": 10.562817573547363,
      "learning_rate": 0.0,
      "loss": 0.4004,
      "step": 1800
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.6403162055335968,
      "eval_loss": 0.8710843324661255,
      "eval_runtime": 28.7285,
      "eval_samples_per_second": 35.226,
      "eval_steps_per_second": 0.557,
      "step": 1800
    },
    {
      "epoch": 20.0,
      "step": 1800,
      "total_flos": 0.0,
      "train_loss": 0.6163041236665514,
      "train_runtime": 10385.8394,
      "train_samples_per_second": 11.034,
      "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
}