{
  "best_metric": 0.8523960709571838,
  "best_model_checkpoint": "./kd_results/apple/mobilevit-small_alpha0.7_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": 5.101388931274414,
      "learning_rate": 6.944444444444445e-06,
      "loss": 1.3523,
      "step": 25
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 2.054196834564209,
      "learning_rate": 1.388888888888889e-05,
      "loss": 1.1546,
      "step": 50
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 2.0043528079986572,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 1.1499,
      "step": 75
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.24604743083003952,
      "eval_loss": 1.4250729084014893,
      "eval_runtime": 29.5953,
      "eval_samples_per_second": 34.195,
      "eval_steps_per_second": 0.541,
      "step": 90
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 2.849222421646118,
      "learning_rate": 2.777777777777778e-05,
      "loss": 1.1279,
      "step": 100
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 3.0925490856170654,
      "learning_rate": 3.472222222222222e-05,
      "loss": 1.1186,
      "step": 125
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 1.8983948230743408,
      "learning_rate": 4.166666666666667e-05,
      "loss": 1.1384,
      "step": 150
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 3.4966299533843994,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 1.0997,
      "step": 175
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.325098814229249,
      "eval_loss": 1.338444709777832,
      "eval_runtime": 28.9738,
      "eval_samples_per_second": 34.928,
      "eval_steps_per_second": 0.552,
      "step": 180
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 5.060726642608643,
      "learning_rate": 4.938271604938271e-05,
      "loss": 1.0582,
      "step": 200
    },
    {
      "epoch": 2.5,
      "grad_norm": 4.789320468902588,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 1.0387,
      "step": 225
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 4.626663684844971,
      "learning_rate": 4.783950617283951e-05,
      "loss": 1.0075,
      "step": 250
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.38636363636363635,
      "eval_loss": 1.2409471273422241,
      "eval_runtime": 28.1993,
      "eval_samples_per_second": 35.887,
      "eval_steps_per_second": 0.567,
      "step": 270
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 4.179274559020996,
      "learning_rate": 4.70679012345679e-05,
      "loss": 0.9792,
      "step": 275
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 4.7588958740234375,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.9014,
      "step": 300
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 4.043580055236816,
      "learning_rate": 4.5524691358024696e-05,
      "loss": 0.8837,
      "step": 325
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 3.6913771629333496,
      "learning_rate": 4.4753086419753084e-05,
      "loss": 0.8689,
      "step": 350
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.4980237154150198,
      "eval_loss": 1.0919088125228882,
      "eval_runtime": 28.6898,
      "eval_samples_per_second": 35.274,
      "eval_steps_per_second": 0.558,
      "step": 360
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 4.968973636627197,
      "learning_rate": 4.3981481481481486e-05,
      "loss": 0.8125,
      "step": 375
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 5.1979169845581055,
      "learning_rate": 4.3209876543209875e-05,
      "loss": 0.7491,
      "step": 400
    },
    {
      "epoch": 4.722222222222222,
      "grad_norm": 5.4226579666137695,
      "learning_rate": 4.243827160493827e-05,
      "loss": 0.7336,
      "step": 425
    },
    {
      "epoch": 5.0,
      "grad_norm": 6.372869968414307,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.7581,
      "step": 450
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5652173913043478,
      "eval_loss": 1.0005403757095337,
      "eval_runtime": 28.5341,
      "eval_samples_per_second": 35.466,
      "eval_steps_per_second": 0.561,
      "step": 450
    },
    {
      "epoch": 5.277777777777778,
      "grad_norm": 4.292941093444824,
      "learning_rate": 4.089506172839506e-05,
      "loss": 0.7154,
      "step": 475
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 8.582777976989746,
      "learning_rate": 4.012345679012346e-05,
      "loss": 0.7083,
      "step": 500
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 5.746469497680664,
      "learning_rate": 3.935185185185186e-05,
      "loss": 0.7002,
      "step": 525
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6037549407114624,
      "eval_loss": 0.9315200448036194,
      "eval_runtime": 28.0649,
      "eval_samples_per_second": 36.059,
      "eval_steps_per_second": 0.57,
      "step": 540
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 4.643389701843262,
      "learning_rate": 3.8580246913580246e-05,
      "loss": 0.6493,
      "step": 550
    },
    {
      "epoch": 6.388888888888889,
      "grad_norm": 6.32736349105835,
      "learning_rate": 3.780864197530865e-05,
      "loss": 0.6441,
      "step": 575
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 9.691339492797852,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.6358,
      "step": 600
    },
    {
      "epoch": 6.944444444444445,
      "grad_norm": 4.686661720275879,
      "learning_rate": 3.626543209876543e-05,
      "loss": 0.6484,
      "step": 625
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.616600790513834,
      "eval_loss": 0.9004554152488708,
      "eval_runtime": 28.2265,
      "eval_samples_per_second": 35.853,
      "eval_steps_per_second": 0.567,
      "step": 630
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 6.596512794494629,
      "learning_rate": 3.5493827160493834e-05,
      "loss": 0.5927,
      "step": 650
    },
    {
      "epoch": 7.5,
      "grad_norm": 8.201225280761719,
      "learning_rate": 3.472222222222222e-05,
      "loss": 0.5884,
      "step": 675
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 5.975377082824707,
      "learning_rate": 3.395061728395062e-05,
      "loss": 0.6085,
      "step": 700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.5988142292490118,
      "eval_loss": 0.9251823425292969,
      "eval_runtime": 27.7044,
      "eval_samples_per_second": 36.529,
      "eval_steps_per_second": 0.578,
      "step": 720
    },
    {
      "epoch": 8.055555555555555,
      "grad_norm": 4.984492301940918,
      "learning_rate": 3.317901234567901e-05,
      "loss": 0.595,
      "step": 725
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 5.812857151031494,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.5634,
      "step": 750
    },
    {
      "epoch": 8.61111111111111,
      "grad_norm": 4.561792850494385,
      "learning_rate": 3.16358024691358e-05,
      "loss": 0.5634,
      "step": 775
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 8.128091812133789,
      "learning_rate": 3.08641975308642e-05,
      "loss": 0.5562,
      "step": 800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6630434782608695,
      "eval_loss": 0.8604928255081177,
      "eval_runtime": 27.8272,
      "eval_samples_per_second": 36.367,
      "eval_steps_per_second": 0.575,
      "step": 810
    },
    {
      "epoch": 9.166666666666666,
      "grad_norm": 6.267117023468018,
      "learning_rate": 3.0092592592592593e-05,
      "loss": 0.5443,
      "step": 825
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 4.6019697189331055,
      "learning_rate": 2.9320987654320992e-05,
      "loss": 0.5428,
      "step": 850
    },
    {
      "epoch": 9.722222222222221,
      "grad_norm": 10.452680587768555,
      "learning_rate": 2.8549382716049384e-05,
      "loss": 0.5432,
      "step": 875
    },
    {
      "epoch": 10.0,
      "grad_norm": 9.50076961517334,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.5352,
      "step": 900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.6264822134387352,
      "eval_loss": 0.8695505857467651,
      "eval_runtime": 27.8937,
      "eval_samples_per_second": 36.281,
      "eval_steps_per_second": 0.574,
      "step": 900
    },
    {
      "epoch": 10.277777777777779,
      "grad_norm": 2.5697309970855713,
      "learning_rate": 2.700617283950617e-05,
      "loss": 0.5147,
      "step": 925
    },
    {
      "epoch": 10.555555555555555,
      "grad_norm": 5.974382400512695,
      "learning_rate": 2.623456790123457e-05,
      "loss": 0.5114,
      "step": 950
    },
    {
      "epoch": 10.833333333333334,
      "grad_norm": 8.94517707824707,
      "learning_rate": 2.5462962962962965e-05,
      "loss": 0.514,
      "step": 975
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.6571146245059288,
      "eval_loss": 0.8556633591651917,
      "eval_runtime": 28.4144,
      "eval_samples_per_second": 35.616,
      "eval_steps_per_second": 0.563,
      "step": 990
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 6.643627643585205,
      "learning_rate": 2.4691358024691357e-05,
      "loss": 0.5034,
      "step": 1000
    },
    {
      "epoch": 11.38888888888889,
      "grad_norm": 5.745645046234131,
      "learning_rate": 2.3919753086419755e-05,
      "loss": 0.4805,
      "step": 1025
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 6.034335613250732,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.4911,
      "step": 1050
    },
    {
      "epoch": 11.944444444444445,
      "grad_norm": 3.848392963409424,
      "learning_rate": 2.2376543209876542e-05,
      "loss": 0.4819,
      "step": 1075
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.6294466403162056,
      "eval_loss": 0.8792505264282227,
      "eval_runtime": 28.0777,
      "eval_samples_per_second": 36.043,
      "eval_steps_per_second": 0.57,
      "step": 1080
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 2.8963096141815186,
      "learning_rate": 2.1604938271604937e-05,
      "loss": 0.4663,
      "step": 1100
    },
    {
      "epoch": 12.5,
      "grad_norm": 5.107327938079834,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.4576,
      "step": 1125
    },
    {
      "epoch": 12.777777777777779,
      "grad_norm": 4.26367712020874,
      "learning_rate": 2.006172839506173e-05,
      "loss": 0.4538,
      "step": 1150
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.6353754940711462,
      "eval_loss": 0.8856125473976135,
      "eval_runtime": 28.3855,
      "eval_samples_per_second": 35.652,
      "eval_steps_per_second": 0.564,
      "step": 1170
    },
    {
      "epoch": 13.055555555555555,
      "grad_norm": 5.3810930252075195,
      "learning_rate": 1.9290123456790123e-05,
      "loss": 0.4551,
      "step": 1175
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 6.8727312088012695,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.4504,
      "step": 1200
    },
    {
      "epoch": 13.61111111111111,
      "grad_norm": 7.505817890167236,
      "learning_rate": 1.7746913580246917e-05,
      "loss": 0.4407,
      "step": 1225
    },
    {
      "epoch": 13.88888888888889,
      "grad_norm": 5.269619941711426,
      "learning_rate": 1.697530864197531e-05,
      "loss": 0.4382,
      "step": 1250
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.658102766798419,
      "eval_loss": 0.8553591966629028,
      "eval_runtime": 28.5817,
      "eval_samples_per_second": 35.407,
      "eval_steps_per_second": 0.56,
      "step": 1260
    },
    {
      "epoch": 14.166666666666666,
      "grad_norm": 4.635805130004883,
      "learning_rate": 1.6203703703703704e-05,
      "loss": 0.4489,
      "step": 1275
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 3.415487051010132,
      "learning_rate": 1.54320987654321e-05,
      "loss": 0.4232,
      "step": 1300
    },
    {
      "epoch": 14.722222222222221,
      "grad_norm": 5.987606048583984,
      "learning_rate": 1.4660493827160496e-05,
      "loss": 0.4332,
      "step": 1325
    },
    {
      "epoch": 15.0,
      "grad_norm": 10.427629470825195,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.4349,
      "step": 1350
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.6393280632411067,
      "eval_loss": 0.8707471489906311,
      "eval_runtime": 28.6276,
      "eval_samples_per_second": 35.35,
      "eval_steps_per_second": 0.559,
      "step": 1350
    },
    {
      "epoch": 15.277777777777779,
      "grad_norm": 4.101249694824219,
      "learning_rate": 1.3117283950617285e-05,
      "loss": 0.415,
      "step": 1375
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 5.26914119720459,
      "learning_rate": 1.2345679012345678e-05,
      "loss": 0.4307,
      "step": 1400
    },
    {
      "epoch": 15.833333333333334,
      "grad_norm": 5.636514186859131,
      "learning_rate": 1.1574074074074075e-05,
      "loss": 0.4122,
      "step": 1425
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.6640316205533597,
      "eval_loss": 0.856772780418396,
      "eval_runtime": 28.5705,
      "eval_samples_per_second": 35.421,
      "eval_steps_per_second": 0.56,
      "step": 1440
    },
    {
      "epoch": 16.11111111111111,
      "grad_norm": 2.9400148391723633,
      "learning_rate": 1.0802469135802469e-05,
      "loss": 0.4016,
      "step": 1450
    },
    {
      "epoch": 16.38888888888889,
      "grad_norm": 5.095480442047119,
      "learning_rate": 1.0030864197530866e-05,
      "loss": 0.4161,
      "step": 1475
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 6.343998908996582,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.4148,
      "step": 1500
    },
    {
      "epoch": 16.944444444444443,
      "grad_norm": 5.17103910446167,
      "learning_rate": 8.487654320987654e-06,
      "loss": 0.4101,
      "step": 1525
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.6630434782608695,
      "eval_loss": 0.8540449142456055,
      "eval_runtime": 28.2778,
      "eval_samples_per_second": 35.788,
      "eval_steps_per_second": 0.566,
      "step": 1530
    },
    {
      "epoch": 17.22222222222222,
      "grad_norm": 6.506363868713379,
      "learning_rate": 7.71604938271605e-06,
      "loss": 0.403,
      "step": 1550
    },
    {
      "epoch": 17.5,
      "grad_norm": 4.386175155639648,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.3955,
      "step": 1575
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 3.3834517002105713,
      "learning_rate": 6.172839506172839e-06,
      "loss": 0.3922,
      "step": 1600
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.6620553359683794,
      "eval_loss": 0.854274332523346,
      "eval_runtime": 28.3553,
      "eval_samples_per_second": 35.69,
      "eval_steps_per_second": 0.564,
      "step": 1620
    },
    {
      "epoch": 18.055555555555557,
      "grad_norm": 3.2310118675231934,
      "learning_rate": 5.401234567901234e-06,
      "loss": 0.4028,
      "step": 1625
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 4.10606050491333,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.3973,
      "step": 1650
    },
    {
      "epoch": 18.61111111111111,
      "grad_norm": 4.001497268676758,
      "learning_rate": 3.858024691358025e-06,
      "loss": 0.3924,
      "step": 1675
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 3.569098472595215,
      "learning_rate": 3.0864197530864196e-06,
      "loss": 0.3929,
      "step": 1700
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.66600790513834,
      "eval_loss": 0.8523960709571838,
      "eval_runtime": 28.5416,
      "eval_samples_per_second": 35.457,
      "eval_steps_per_second": 0.561,
      "step": 1710
    },
    {
      "epoch": 19.166666666666668,
      "grad_norm": 3.8873579502105713,
      "learning_rate": 2.3148148148148148e-06,
      "loss": 0.3985,
      "step": 1725
    },
    {
      "epoch": 19.444444444444443,
      "grad_norm": 2.5558040142059326,
      "learning_rate": 1.5432098765432098e-06,
      "loss": 0.3918,
      "step": 1750
    },
    {
      "epoch": 19.72222222222222,
      "grad_norm": 5.380621433258057,
      "learning_rate": 7.716049382716049e-07,
      "loss": 0.3961,
      "step": 1775
    },
    {
      "epoch": 20.0,
      "grad_norm": 10.76089096069336,
      "learning_rate": 0.0,
      "loss": 0.3965,
      "step": 1800
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.6511857707509882,
      "eval_loss": 0.8660988807678223,
      "eval_runtime": 28.6261,
      "eval_samples_per_second": 35.352,
      "eval_steps_per_second": 0.559,
      "step": 1800
    },
    {
      "epoch": 20.0,
      "step": 1800,
      "total_flos": 0.0,
      "train_loss": 0.6184117730458577,
      "train_runtime": 10375.0323,
      "train_samples_per_second": 11.046,
      "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
}