{
  "best_metric": 0.5564627051353455,
  "best_model_checkpoint": "./kd_results/microsoft/beit-base-patch16-224_alpha0.7_temp3.0/checkpoint-1170",
  "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": 5.002269744873047,
      "learning_rate": 6.944444444444445e-06,
      "loss": 1.2175,
      "step": 25
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 4.198030948638916,
      "learning_rate": 1.388888888888889e-05,
      "loss": 1.1526,
      "step": 50
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 7.540940284729004,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 1.1373,
      "step": 75
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.28063241106719367,
      "eval_loss": 1.4267971515655518,
      "eval_runtime": 31.08,
      "eval_samples_per_second": 32.561,
      "eval_steps_per_second": 0.515,
      "step": 90
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 5.299309730529785,
      "learning_rate": 2.777777777777778e-05,
      "loss": 1.0902,
      "step": 100
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 3.794218063354492,
      "learning_rate": 3.472222222222222e-05,
      "loss": 1.0292,
      "step": 125
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 6.296744346618652,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.9742,
      "step": 150
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 6.439919471740723,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 0.902,
      "step": 175
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.532608695652174,
      "eval_loss": 1.1061384677886963,
      "eval_runtime": 30.0137,
      "eval_samples_per_second": 33.718,
      "eval_steps_per_second": 0.533,
      "step": 180
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 7.4073662757873535,
      "learning_rate": 4.938271604938271e-05,
      "loss": 0.7105,
      "step": 200
    },
    {
      "epoch": 2.5,
      "grad_norm": 6.429058074951172,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 0.6478,
      "step": 225
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 5.277145862579346,
      "learning_rate": 4.783950617283951e-05,
      "loss": 0.5748,
      "step": 250
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7401185770750988,
      "eval_loss": 0.7085344791412354,
      "eval_runtime": 30.1131,
      "eval_samples_per_second": 33.607,
      "eval_steps_per_second": 0.531,
      "step": 270
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 4.95331335067749,
      "learning_rate": 4.70679012345679e-05,
      "loss": 0.5046,
      "step": 275
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 3.85587215423584,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.4034,
      "step": 300
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 5.835513591766357,
      "learning_rate": 4.5524691358024696e-05,
      "loss": 0.3894,
      "step": 325
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 5.092733383178711,
      "learning_rate": 4.4753086419753084e-05,
      "loss": 0.3815,
      "step": 350
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7687747035573123,
      "eval_loss": 0.6585662364959717,
      "eval_runtime": 30.7477,
      "eval_samples_per_second": 32.913,
      "eval_steps_per_second": 0.52,
      "step": 360
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 6.647624969482422,
      "learning_rate": 4.3981481481481486e-05,
      "loss": 0.3212,
      "step": 375
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 3.9873390197753906,
      "learning_rate": 4.3209876543209875e-05,
      "loss": 0.2895,
      "step": 400
    },
    {
      "epoch": 4.722222222222222,
      "grad_norm": 3.4268240928649902,
      "learning_rate": 4.243827160493827e-05,
      "loss": 0.273,
      "step": 425
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.725526809692383,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.2747,
      "step": 450
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7924901185770751,
      "eval_loss": 0.597775399684906,
      "eval_runtime": 31.7866,
      "eval_samples_per_second": 31.837,
      "eval_steps_per_second": 0.503,
      "step": 450
    },
    {
      "epoch": 5.277777777777778,
      "grad_norm": 3.5042800903320312,
      "learning_rate": 4.089506172839506e-05,
      "loss": 0.2418,
      "step": 475
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 3.098809242248535,
      "learning_rate": 4.012345679012346e-05,
      "loss": 0.2205,
      "step": 500
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 2.3750343322753906,
      "learning_rate": 3.935185185185186e-05,
      "loss": 0.2098,
      "step": 525
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7895256916996047,
      "eval_loss": 0.6237751841545105,
      "eval_runtime": 30.2211,
      "eval_samples_per_second": 33.487,
      "eval_steps_per_second": 0.529,
      "step": 540
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 2.2495975494384766,
      "learning_rate": 3.8580246913580246e-05,
      "loss": 0.2133,
      "step": 550
    },
    {
      "epoch": 6.388888888888889,
      "grad_norm": 2.4142961502075195,
      "learning_rate": 3.780864197530865e-05,
      "loss": 0.2076,
      "step": 575
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 2.8467612266540527,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.2012,
      "step": 600
    },
    {
      "epoch": 6.944444444444445,
      "grad_norm": 2.975644826889038,
      "learning_rate": 3.626543209876543e-05,
      "loss": 0.1954,
      "step": 625
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7747035573122529,
      "eval_loss": 0.6208274364471436,
      "eval_runtime": 30.4524,
      "eval_samples_per_second": 33.232,
      "eval_steps_per_second": 0.525,
      "step": 630
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 3.393961191177368,
      "learning_rate": 3.5493827160493834e-05,
      "loss": 0.1797,
      "step": 650
    },
    {
      "epoch": 7.5,
      "grad_norm": 1.9453306198120117,
      "learning_rate": 3.472222222222222e-05,
      "loss": 0.179,
      "step": 675
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 1.9624695777893066,
      "learning_rate": 3.395061728395062e-05,
      "loss": 0.182,
      "step": 700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8162055335968379,
      "eval_loss": 0.5784512758255005,
      "eval_runtime": 30.5488,
      "eval_samples_per_second": 33.127,
      "eval_steps_per_second": 0.524,
      "step": 720
    },
    {
      "epoch": 8.055555555555555,
      "grad_norm": 2.131540298461914,
      "learning_rate": 3.317901234567901e-05,
      "loss": 0.1814,
      "step": 725
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 1.745793104171753,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.1742,
      "step": 750
    },
    {
      "epoch": 8.61111111111111,
      "grad_norm": 1.6976990699768066,
      "learning_rate": 3.16358024691358e-05,
      "loss": 0.1661,
      "step": 775
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 2.869600534439087,
      "learning_rate": 3.08641975308642e-05,
      "loss": 0.1748,
      "step": 800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.799407114624506,
      "eval_loss": 0.57688307762146,
      "eval_runtime": 31.6764,
      "eval_samples_per_second": 31.948,
      "eval_steps_per_second": 0.505,
      "step": 810
    },
    {
      "epoch": 9.166666666666666,
      "grad_norm": 2.1438984870910645,
      "learning_rate": 3.0092592592592593e-05,
      "loss": 0.1609,
      "step": 825
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 1.322683334350586,
      "learning_rate": 2.9320987654320992e-05,
      "loss": 0.1575,
      "step": 850
    },
    {
      "epoch": 9.722222222222221,
      "grad_norm": 1.389501929283142,
      "learning_rate": 2.8549382716049384e-05,
      "loss": 0.1584,
      "step": 875
    },
    {
      "epoch": 10.0,
      "grad_norm": 2.022207498550415,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.1568,
      "step": 900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.808300395256917,
      "eval_loss": 0.5669097304344177,
      "eval_runtime": 30.7621,
      "eval_samples_per_second": 32.898,
      "eval_steps_per_second": 0.52,
      "step": 900
    },
    {
      "epoch": 10.277777777777779,
      "grad_norm": 1.6512681245803833,
      "learning_rate": 2.700617283950617e-05,
      "loss": 0.1596,
      "step": 925
    },
    {
      "epoch": 10.555555555555555,
      "grad_norm": 1.49599027633667,
      "learning_rate": 2.623456790123457e-05,
      "loss": 0.1521,
      "step": 950
    },
    {
      "epoch": 10.833333333333334,
      "grad_norm": 1.6968499422073364,
      "learning_rate": 2.5462962962962965e-05,
      "loss": 0.1542,
      "step": 975
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.799407114624506,
      "eval_loss": 0.5793797373771667,
      "eval_runtime": 30.4106,
      "eval_samples_per_second": 33.278,
      "eval_steps_per_second": 0.526,
      "step": 990
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 1.3159855604171753,
      "learning_rate": 2.4691358024691357e-05,
      "loss": 0.1436,
      "step": 1000
    },
    {
      "epoch": 11.38888888888889,
      "grad_norm": 1.3499946594238281,
      "learning_rate": 2.3919753086419755e-05,
      "loss": 0.1369,
      "step": 1025
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 1.3021754026412964,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.1472,
      "step": 1050
    },
    {
      "epoch": 11.944444444444445,
      "grad_norm": 0.9485201835632324,
      "learning_rate": 2.2376543209876542e-05,
      "loss": 0.1417,
      "step": 1075
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.7934782608695652,
      "eval_loss": 0.5789390802383423,
      "eval_runtime": 30.8856,
      "eval_samples_per_second": 32.766,
      "eval_steps_per_second": 0.518,
      "step": 1080
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 1.0141340494155884,
      "learning_rate": 2.1604938271604937e-05,
      "loss": 0.1367,
      "step": 1100
    },
    {
      "epoch": 12.5,
      "grad_norm": 1.3728690147399902,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.1471,
      "step": 1125
    },
    {
      "epoch": 12.777777777777779,
      "grad_norm": 1.656044602394104,
      "learning_rate": 2.006172839506173e-05,
      "loss": 0.1335,
      "step": 1150
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.8142292490118577,
      "eval_loss": 0.5564627051353455,
      "eval_runtime": 31.6838,
      "eval_samples_per_second": 31.941,
      "eval_steps_per_second": 0.505,
      "step": 1170
    },
    {
      "epoch": 13.055555555555555,
      "grad_norm": 1.058284878730774,
      "learning_rate": 1.9290123456790123e-05,
      "loss": 0.1299,
      "step": 1175
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 1.1719366312026978,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.1374,
      "step": 1200
    },
    {
      "epoch": 13.61111111111111,
      "grad_norm": 0.807572066783905,
      "learning_rate": 1.7746913580246917e-05,
      "loss": 0.1322,
      "step": 1225
    },
    {
      "epoch": 13.88888888888889,
      "grad_norm": 0.76854407787323,
      "learning_rate": 1.697530864197531e-05,
      "loss": 0.126,
      "step": 1250
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8162055335968379,
      "eval_loss": 0.5735021233558655,
      "eval_runtime": 30.8997,
      "eval_samples_per_second": 32.751,
      "eval_steps_per_second": 0.518,
      "step": 1260
    },
    {
      "epoch": 14.166666666666666,
      "grad_norm": 1.64256751537323,
      "learning_rate": 1.6203703703703704e-05,
      "loss": 0.1312,
      "step": 1275
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 1.0654799938201904,
      "learning_rate": 1.54320987654321e-05,
      "loss": 0.1288,
      "step": 1300
    },
    {
      "epoch": 14.722222222222221,
      "grad_norm": 1.1760451793670654,
      "learning_rate": 1.4660493827160496e-05,
      "loss": 0.1246,
      "step": 1325
    },
    {
      "epoch": 15.0,
      "grad_norm": 1.2862129211425781,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.1277,
      "step": 1350
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.7964426877470355,
      "eval_loss": 0.5776201486587524,
      "eval_runtime": 31.0336,
      "eval_samples_per_second": 32.61,
      "eval_steps_per_second": 0.516,
      "step": 1350
    },
    {
      "epoch": 15.277777777777779,
      "grad_norm": 0.8567253947257996,
      "learning_rate": 1.3117283950617285e-05,
      "loss": 0.1271,
      "step": 1375
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 1.2438898086547852,
      "learning_rate": 1.2345679012345678e-05,
      "loss": 0.1229,
      "step": 1400
    },
    {
      "epoch": 15.833333333333334,
      "grad_norm": 1.0080410242080688,
      "learning_rate": 1.1574074074074075e-05,
      "loss": 0.1226,
      "step": 1425
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.8033596837944664,
      "eval_loss": 0.559334397315979,
      "eval_runtime": 31.2467,
      "eval_samples_per_second": 32.387,
      "eval_steps_per_second": 0.512,
      "step": 1440
    },
    {
      "epoch": 16.11111111111111,
      "grad_norm": 0.6350163221359253,
      "learning_rate": 1.0802469135802469e-05,
      "loss": 0.1214,
      "step": 1450
    },
    {
      "epoch": 16.38888888888889,
      "grad_norm": 0.7633430361747742,
      "learning_rate": 1.0030864197530866e-05,
      "loss": 0.1185,
      "step": 1475
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 0.7166378498077393,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.1226,
      "step": 1500
    },
    {
      "epoch": 16.944444444444443,
      "grad_norm": 0.8270022869110107,
      "learning_rate": 8.487654320987654e-06,
      "loss": 0.1238,
      "step": 1525
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.808300395256917,
      "eval_loss": 0.5681459903717041,
      "eval_runtime": 32.0583,
      "eval_samples_per_second": 31.568,
      "eval_steps_per_second": 0.499,
      "step": 1530
    },
    {
      "epoch": 17.22222222222222,
      "grad_norm": 0.7186818718910217,
      "learning_rate": 7.71604938271605e-06,
      "loss": 0.1211,
      "step": 1550
    },
    {
      "epoch": 17.5,
      "grad_norm": 0.7400338053703308,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.1184,
      "step": 1575
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 1.0720752477645874,
      "learning_rate": 6.172839506172839e-06,
      "loss": 0.1149,
      "step": 1600
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.8063241106719368,
      "eval_loss": 0.5669211745262146,
      "eval_runtime": 31.0414,
      "eval_samples_per_second": 32.602,
      "eval_steps_per_second": 0.515,
      "step": 1620
    },
    {
      "epoch": 18.055555555555557,
      "grad_norm": 0.8105254173278809,
      "learning_rate": 5.401234567901234e-06,
      "loss": 0.1224,
      "step": 1625
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 0.6114757657051086,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.1207,
      "step": 1650
    },
    {
      "epoch": 18.61111111111111,
      "grad_norm": 0.8441098928451538,
      "learning_rate": 3.858024691358025e-06,
      "loss": 0.1125,
      "step": 1675
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 0.5961729884147644,
      "learning_rate": 3.0864197530864196e-06,
      "loss": 0.1156,
      "step": 1700
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.8102766798418972,
      "eval_loss": 0.5672557950019836,
      "eval_runtime": 31.1092,
      "eval_samples_per_second": 32.531,
      "eval_steps_per_second": 0.514,
      "step": 1710
    },
    {
      "epoch": 19.166666666666668,
      "grad_norm": 0.521712601184845,
      "learning_rate": 2.3148148148148148e-06,
      "loss": 0.1138,
      "step": 1725
    },
    {
      "epoch": 19.444444444444443,
      "grad_norm": 0.5970616340637207,
      "learning_rate": 1.5432098765432098e-06,
      "loss": 0.1186,
      "step": 1750
    },
    {
      "epoch": 19.72222222222222,
      "grad_norm": 0.6230306625366211,
      "learning_rate": 7.716049382716049e-07,
      "loss": 0.1163,
      "step": 1775
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.8082752227783203,
      "learning_rate": 0.0,
      "loss": 0.1127,
      "step": 1800
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8102766798418972,
      "eval_loss": 0.5651406049728394,
      "eval_runtime": 31.6271,
      "eval_samples_per_second": 31.998,
      "eval_steps_per_second": 0.506,
      "step": 1800
    },
    {
      "epoch": 20.0,
      "step": 1800,
      "total_flos": 0.0,
      "train_loss": 0.28153025918536717,
      "train_runtime": 10921.7049,
      "train_samples_per_second": 10.493,
      "train_steps_per_second": 0.165
    }
  ],
  "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
}