{
  "best_metric": 0.8373316526412964,
  "best_model_checkpoint": "./kd_results/apple/mobilevit-x-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": 3.247164249420166,
      "learning_rate": 6.944444444444445e-06,
      "loss": 1.2066,
      "step": 25
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 2.215648889541626,
      "learning_rate": 1.388888888888889e-05,
      "loss": 1.1255,
      "step": 50
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 2.30334210395813,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 1.1201,
      "step": 75
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.29150197628458496,
      "eval_loss": 1.390669822692871,
      "eval_runtime": 29.6191,
      "eval_samples_per_second": 34.167,
      "eval_steps_per_second": 0.54,
      "step": 90
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 2.3844776153564453,
      "learning_rate": 2.777777777777778e-05,
      "loss": 1.1054,
      "step": 100
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 2.0609209537506104,
      "learning_rate": 3.472222222222222e-05,
      "loss": 1.0973,
      "step": 125
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 1.205605387687683,
      "learning_rate": 4.166666666666667e-05,
      "loss": 1.1245,
      "step": 150
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 2.681898355484009,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 1.0793,
      "step": 175
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3774703557312253,
      "eval_loss": 1.3214282989501953,
      "eval_runtime": 29.1865,
      "eval_samples_per_second": 34.674,
      "eval_steps_per_second": 0.548,
      "step": 180
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 4.117778778076172,
      "learning_rate": 4.938271604938271e-05,
      "loss": 1.028,
      "step": 200
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.7841434478759766,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 1.0231,
      "step": 225
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 2.144334554672241,
      "learning_rate": 4.783950617283951e-05,
      "loss": 0.9679,
      "step": 250
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4644268774703557,
      "eval_loss": 1.176648497581482,
      "eval_runtime": 28.4142,
      "eval_samples_per_second": 35.616,
      "eval_steps_per_second": 0.563,
      "step": 270
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 3.034731149673462,
      "learning_rate": 4.70679012345679e-05,
      "loss": 0.9522,
      "step": 275
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 3.2194249629974365,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.8819,
      "step": 300
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 4.528558731079102,
      "learning_rate": 4.5524691358024696e-05,
      "loss": 0.8617,
      "step": 325
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 2.845184564590454,
      "learning_rate": 4.4753086419753084e-05,
      "loss": 0.844,
      "step": 350
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5316205533596838,
      "eval_loss": 1.0510571002960205,
      "eval_runtime": 28.5015,
      "eval_samples_per_second": 35.507,
      "eval_steps_per_second": 0.561,
      "step": 360
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 4.721632957458496,
      "learning_rate": 4.3981481481481486e-05,
      "loss": 0.8004,
      "step": 375
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 4.212678909301758,
      "learning_rate": 4.3209876543209875e-05,
      "loss": 0.748,
      "step": 400
    },
    {
      "epoch": 4.722222222222222,
      "grad_norm": 3.4344043731689453,
      "learning_rate": 4.243827160493827e-05,
      "loss": 0.7596,
      "step": 425
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.402198314666748,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.745,
      "step": 450
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5810276679841897,
      "eval_loss": 0.9648578763008118,
      "eval_runtime": 29.0551,
      "eval_samples_per_second": 34.83,
      "eval_steps_per_second": 0.551,
      "step": 450
    },
    {
      "epoch": 5.277777777777778,
      "grad_norm": 2.8925297260284424,
      "learning_rate": 4.089506172839506e-05,
      "loss": 0.7187,
      "step": 475
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 5.771768093109131,
      "learning_rate": 4.012345679012346e-05,
      "loss": 0.718,
      "step": 500
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 3.5513947010040283,
      "learning_rate": 3.935185185185186e-05,
      "loss": 0.7137,
      "step": 525
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5800395256916996,
      "eval_loss": 0.9291349649429321,
      "eval_runtime": 29.2524,
      "eval_samples_per_second": 34.595,
      "eval_steps_per_second": 0.547,
      "step": 540
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 3.8858561515808105,
      "learning_rate": 3.8580246913580246e-05,
      "loss": 0.6701,
      "step": 550
    },
    {
      "epoch": 6.388888888888889,
      "grad_norm": 4.007695198059082,
      "learning_rate": 3.780864197530865e-05,
      "loss": 0.6809,
      "step": 575
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 3.8885931968688965,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.6539,
      "step": 600
    },
    {
      "epoch": 6.944444444444445,
      "grad_norm": 4.412193775177002,
      "learning_rate": 3.626543209876543e-05,
      "loss": 0.6625,
      "step": 625
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6235177865612648,
      "eval_loss": 0.8924146890640259,
      "eval_runtime": 28.9801,
      "eval_samples_per_second": 34.92,
      "eval_steps_per_second": 0.552,
      "step": 630
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 4.455688953399658,
      "learning_rate": 3.5493827160493834e-05,
      "loss": 0.6276,
      "step": 650
    },
    {
      "epoch": 7.5,
      "grad_norm": 4.672175884246826,
      "learning_rate": 3.472222222222222e-05,
      "loss": 0.641,
      "step": 675
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 6.414026737213135,
      "learning_rate": 3.395061728395062e-05,
      "loss": 0.6515,
      "step": 700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.6017786561264822,
      "eval_loss": 0.9112664461135864,
      "eval_runtime": 29.2789,
      "eval_samples_per_second": 34.564,
      "eval_steps_per_second": 0.546,
      "step": 720
    },
    {
      "epoch": 8.055555555555555,
      "grad_norm": 4.316091537475586,
      "learning_rate": 3.317901234567901e-05,
      "loss": 0.6347,
      "step": 725
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 5.868544578552246,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.6113,
      "step": 750
    },
    {
      "epoch": 8.61111111111111,
      "grad_norm": 3.766305923461914,
      "learning_rate": 3.16358024691358e-05,
      "loss": 0.6321,
      "step": 775
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 3.8639538288116455,
      "learning_rate": 3.08641975308642e-05,
      "loss": 0.604,
      "step": 800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6324110671936759,
      "eval_loss": 0.8726031184196472,
      "eval_runtime": 28.859,
      "eval_samples_per_second": 35.067,
      "eval_steps_per_second": 0.554,
      "step": 810
    },
    {
      "epoch": 9.166666666666666,
      "grad_norm": 4.397319793701172,
      "learning_rate": 3.0092592592592593e-05,
      "loss": 0.5992,
      "step": 825
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 4.5495686531066895,
      "learning_rate": 2.9320987654320992e-05,
      "loss": 0.5977,
      "step": 850
    },
    {
      "epoch": 9.722222222222221,
      "grad_norm": 6.600956916809082,
      "learning_rate": 2.8549382716049384e-05,
      "loss": 0.5862,
      "step": 875
    },
    {
      "epoch": 10.0,
      "grad_norm": 9.885616302490234,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.6031,
      "step": 900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.6472332015810277,
      "eval_loss": 0.8650755882263184,
      "eval_runtime": 29.0743,
      "eval_samples_per_second": 34.807,
      "eval_steps_per_second": 0.55,
      "step": 900
    },
    {
      "epoch": 10.277777777777779,
      "grad_norm": 5.1479902267456055,
      "learning_rate": 2.700617283950617e-05,
      "loss": 0.5746,
      "step": 925
    },
    {
      "epoch": 10.555555555555555,
      "grad_norm": 5.609706878662109,
      "learning_rate": 2.623456790123457e-05,
      "loss": 0.5669,
      "step": 950
    },
    {
      "epoch": 10.833333333333334,
      "grad_norm": 4.310824394226074,
      "learning_rate": 2.5462962962962965e-05,
      "loss": 0.5679,
      "step": 975
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.6531620553359684,
      "eval_loss": 0.8491111993789673,
      "eval_runtime": 28.6642,
      "eval_samples_per_second": 35.305,
      "eval_steps_per_second": 0.558,
      "step": 990
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 4.360653877258301,
      "learning_rate": 2.4691358024691357e-05,
      "loss": 0.5746,
      "step": 1000
    },
    {
      "epoch": 11.38888888888889,
      "grad_norm": 3.0692925453186035,
      "learning_rate": 2.3919753086419755e-05,
      "loss": 0.5481,
      "step": 1025
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 6.306731700897217,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.5578,
      "step": 1050
    },
    {
      "epoch": 11.944444444444445,
      "grad_norm": 5.40558385848999,
      "learning_rate": 2.2376543209876542e-05,
      "loss": 0.5615,
      "step": 1075
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.6324110671936759,
      "eval_loss": 0.8680435419082642,
      "eval_runtime": 29.0293,
      "eval_samples_per_second": 34.861,
      "eval_steps_per_second": 0.551,
      "step": 1080
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 4.685542583465576,
      "learning_rate": 2.1604938271604937e-05,
      "loss": 0.5421,
      "step": 1100
    },
    {
      "epoch": 12.5,
      "grad_norm": 3.5871431827545166,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.5359,
      "step": 1125
    },
    {
      "epoch": 12.777777777777779,
      "grad_norm": 3.0133259296417236,
      "learning_rate": 2.006172839506173e-05,
      "loss": 0.5364,
      "step": 1150
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.6254940711462451,
      "eval_loss": 0.8812161684036255,
      "eval_runtime": 29.0397,
      "eval_samples_per_second": 34.849,
      "eval_steps_per_second": 0.551,
      "step": 1170
    },
    {
      "epoch": 13.055555555555555,
      "grad_norm": 3.356553077697754,
      "learning_rate": 1.9290123456790123e-05,
      "loss": 0.5271,
      "step": 1175
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 3.72344970703125,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.5299,
      "step": 1200
    },
    {
      "epoch": 13.61111111111111,
      "grad_norm": 6.30882453918457,
      "learning_rate": 1.7746913580246917e-05,
      "loss": 0.5268,
      "step": 1225
    },
    {
      "epoch": 13.88888888888889,
      "grad_norm": 4.017136573791504,
      "learning_rate": 1.697530864197531e-05,
      "loss": 0.507,
      "step": 1250
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.633399209486166,
      "eval_loss": 0.8634254932403564,
      "eval_runtime": 29.3197,
      "eval_samples_per_second": 34.516,
      "eval_steps_per_second": 0.546,
      "step": 1260
    },
    {
      "epoch": 14.166666666666666,
      "grad_norm": 5.9562249183654785,
      "learning_rate": 1.6203703703703704e-05,
      "loss": 0.5189,
      "step": 1275
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 2.8607287406921387,
      "learning_rate": 1.54320987654321e-05,
      "loss": 0.5016,
      "step": 1300
    },
    {
      "epoch": 14.722222222222221,
      "grad_norm": 4.913319110870361,
      "learning_rate": 1.4660493827160496e-05,
      "loss": 0.5076,
      "step": 1325
    },
    {
      "epoch": 15.0,
      "grad_norm": 9.209261894226074,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.5122,
      "step": 1350
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.6383399209486166,
      "eval_loss": 0.872487485408783,
      "eval_runtime": 29.1233,
      "eval_samples_per_second": 34.749,
      "eval_steps_per_second": 0.549,
      "step": 1350
    },
    {
      "epoch": 15.277777777777779,
      "grad_norm": 3.7022054195404053,
      "learning_rate": 1.3117283950617285e-05,
      "loss": 0.4953,
      "step": 1375
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 4.390769958496094,
      "learning_rate": 1.2345679012345678e-05,
      "loss": 0.5101,
      "step": 1400
    },
    {
      "epoch": 15.833333333333334,
      "grad_norm": 5.139510631561279,
      "learning_rate": 1.1574074074074075e-05,
      "loss": 0.501,
      "step": 1425
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.650197628458498,
      "eval_loss": 0.8544252514839172,
      "eval_runtime": 29.2799,
      "eval_samples_per_second": 34.563,
      "eval_steps_per_second": 0.546,
      "step": 1440
    },
    {
      "epoch": 16.11111111111111,
      "grad_norm": 2.6015398502349854,
      "learning_rate": 1.0802469135802469e-05,
      "loss": 0.4846,
      "step": 1450
    },
    {
      "epoch": 16.38888888888889,
      "grad_norm": 5.3646368980407715,
      "learning_rate": 1.0030864197530866e-05,
      "loss": 0.4912,
      "step": 1475
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 5.250799179077148,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.4939,
      "step": 1500
    },
    {
      "epoch": 16.944444444444443,
      "grad_norm": 4.629016876220703,
      "learning_rate": 8.487654320987654e-06,
      "loss": 0.4927,
      "step": 1525
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.650197628458498,
      "eval_loss": 0.8532552123069763,
      "eval_runtime": 28.8516,
      "eval_samples_per_second": 35.076,
      "eval_steps_per_second": 0.555,
      "step": 1530
    },
    {
      "epoch": 17.22222222222222,
      "grad_norm": 3.7632739543914795,
      "learning_rate": 7.71604938271605e-06,
      "loss": 0.4717,
      "step": 1550
    },
    {
      "epoch": 17.5,
      "grad_norm": 3.0579588413238525,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.4886,
      "step": 1575
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 3.2696261405944824,
      "learning_rate": 6.172839506172839e-06,
      "loss": 0.4891,
      "step": 1600
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.6482213438735178,
      "eval_loss": 0.8573007583618164,
      "eval_runtime": 28.6902,
      "eval_samples_per_second": 35.273,
      "eval_steps_per_second": 0.558,
      "step": 1620
    },
    {
      "epoch": 18.055555555555557,
      "grad_norm": 2.90147066116333,
      "learning_rate": 5.401234567901234e-06,
      "loss": 0.4855,
      "step": 1625
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 4.540664196014404,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.4859,
      "step": 1650
    },
    {
      "epoch": 18.61111111111111,
      "grad_norm": 5.820868968963623,
      "learning_rate": 3.858024691358025e-06,
      "loss": 0.4766,
      "step": 1675
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 4.261588096618652,
      "learning_rate": 3.0864197530864196e-06,
      "loss": 0.4767,
      "step": 1700
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.6620553359683794,
      "eval_loss": 0.8373316526412964,
      "eval_runtime": 27.9286,
      "eval_samples_per_second": 36.235,
      "eval_steps_per_second": 0.573,
      "step": 1710
    },
    {
      "epoch": 19.166666666666668,
      "grad_norm": 3.533184051513672,
      "learning_rate": 2.3148148148148148e-06,
      "loss": 0.4921,
      "step": 1725
    },
    {
      "epoch": 19.444444444444443,
      "grad_norm": 3.0779712200164795,
      "learning_rate": 1.5432098765432098e-06,
      "loss": 0.4619,
      "step": 1750
    },
    {
      "epoch": 19.72222222222222,
      "grad_norm": 4.86537504196167,
      "learning_rate": 7.716049382716049e-07,
      "loss": 0.4913,
      "step": 1775
    },
    {
      "epoch": 20.0,
      "grad_norm": 9.412094116210938,
      "learning_rate": 0.0,
      "loss": 0.4903,
      "step": 1800
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.6472332015810277,
      "eval_loss": 0.8546063899993896,
      "eval_runtime": 28.8756,
      "eval_samples_per_second": 35.047,
      "eval_steps_per_second": 0.554,
      "step": 1800
    },
    {
      "epoch": 20.0,
      "step": 1800,
      "total_flos": 0.0,
      "train_loss": 0.6604432890150282,
      "train_runtime": 10425.1813,
      "train_samples_per_second": 10.993,
      "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
}