{
  "best_metric": 0.6259772777557373,
  "best_model_checkpoint": "./kd_results/microsoft/swin-tiny-patch4-window7-224_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": 4.772194862365723,
      "learning_rate": 6.944444444444445e-06,
      "loss": 1.193,
      "step": 25
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 4.768210411071777,
      "learning_rate": 1.388888888888889e-05,
      "loss": 1.1581,
      "step": 50
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 5.157600402832031,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 1.1461,
      "step": 75
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.2816205533596838,
      "eval_loss": 1.4228136539459229,
      "eval_runtime": 29.9921,
      "eval_samples_per_second": 33.742,
      "eval_steps_per_second": 0.533,
      "step": 90
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 6.370566368103027,
      "learning_rate": 2.777777777777778e-05,
      "loss": 1.128,
      "step": 100
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 3.9340157508850098,
      "learning_rate": 3.472222222222222e-05,
      "loss": 1.1074,
      "step": 125
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 4.803001403808594,
      "learning_rate": 4.166666666666667e-05,
      "loss": 1.1224,
      "step": 150
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 6.863511562347412,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 1.0312,
      "step": 175
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4318181818181818,
      "eval_loss": 1.2249794006347656,
      "eval_runtime": 33.3095,
      "eval_samples_per_second": 30.382,
      "eval_steps_per_second": 0.48,
      "step": 180
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 9.351407051086426,
      "learning_rate": 4.938271604938271e-05,
      "loss": 0.905,
      "step": 200
    },
    {
      "epoch": 2.5,
      "grad_norm": 6.47028112411499,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 0.8312,
      "step": 225
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 8.462491035461426,
      "learning_rate": 4.783950617283951e-05,
      "loss": 0.7723,
      "step": 250
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5652173913043478,
      "eval_loss": 1.0165187120437622,
      "eval_runtime": 32.4915,
      "eval_samples_per_second": 31.147,
      "eval_steps_per_second": 0.492,
      "step": 270
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 8.262164115905762,
      "learning_rate": 4.70679012345679e-05,
      "loss": 0.7387,
      "step": 275
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 5.517256736755371,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.5976,
      "step": 300
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 6.088006019592285,
      "learning_rate": 4.5524691358024696e-05,
      "loss": 0.5863,
      "step": 325
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 6.293733596801758,
      "learning_rate": 4.4753086419753084e-05,
      "loss": 0.5804,
      "step": 350
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.708498023715415,
      "eval_loss": 0.7592554092407227,
      "eval_runtime": 32.2869,
      "eval_samples_per_second": 31.344,
      "eval_steps_per_second": 0.496,
      "step": 360
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 7.410552501678467,
      "learning_rate": 4.3981481481481486e-05,
      "loss": 0.4866,
      "step": 375
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 6.7053303718566895,
      "learning_rate": 4.3209876543209875e-05,
      "loss": 0.4433,
      "step": 400
    },
    {
      "epoch": 4.722222222222222,
      "grad_norm": 7.292109966278076,
      "learning_rate": 4.243827160493827e-05,
      "loss": 0.4281,
      "step": 425
    },
    {
      "epoch": 5.0,
      "grad_norm": 7.13463020324707,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.4243,
      "step": 450
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7539525691699605,
      "eval_loss": 0.7048425674438477,
      "eval_runtime": 30.9907,
      "eval_samples_per_second": 32.655,
      "eval_steps_per_second": 0.516,
      "step": 450
    },
    {
      "epoch": 5.277777777777778,
      "grad_norm": 2.790771484375,
      "learning_rate": 4.089506172839506e-05,
      "loss": 0.3519,
      "step": 475
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 4.171605110168457,
      "learning_rate": 4.012345679012346e-05,
      "loss": 0.3263,
      "step": 500
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 4.286325454711914,
      "learning_rate": 3.935185185185186e-05,
      "loss": 0.3253,
      "step": 525
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7608695652173914,
      "eval_loss": 0.6930640935897827,
      "eval_runtime": 31.6106,
      "eval_samples_per_second": 32.015,
      "eval_steps_per_second": 0.506,
      "step": 540
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 4.520972728729248,
      "learning_rate": 3.8580246913580246e-05,
      "loss": 0.3187,
      "step": 550
    },
    {
      "epoch": 6.388888888888889,
      "grad_norm": 3.8681890964508057,
      "learning_rate": 3.780864197530865e-05,
      "loss": 0.2681,
      "step": 575
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 4.683760166168213,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.2803,
      "step": 600
    },
    {
      "epoch": 6.944444444444445,
      "grad_norm": 6.010910511016846,
      "learning_rate": 3.626543209876543e-05,
      "loss": 0.2822,
      "step": 625
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7638339920948617,
      "eval_loss": 0.6610152125358582,
      "eval_runtime": 30.7675,
      "eval_samples_per_second": 32.892,
      "eval_steps_per_second": 0.52,
      "step": 630
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 3.4549858570098877,
      "learning_rate": 3.5493827160493834e-05,
      "loss": 0.2434,
      "step": 650
    },
    {
      "epoch": 7.5,
      "grad_norm": 6.928824424743652,
      "learning_rate": 3.472222222222222e-05,
      "loss": 0.2318,
      "step": 675
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 3.8014657497406006,
      "learning_rate": 3.395061728395062e-05,
      "loss": 0.229,
      "step": 700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.7786561264822134,
      "eval_loss": 0.653826117515564,
      "eval_runtime": 31.3417,
      "eval_samples_per_second": 32.289,
      "eval_steps_per_second": 0.511,
      "step": 720
    },
    {
      "epoch": 8.055555555555555,
      "grad_norm": 2.461010217666626,
      "learning_rate": 3.317901234567901e-05,
      "loss": 0.2305,
      "step": 725
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 3.2131762504577637,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.1993,
      "step": 750
    },
    {
      "epoch": 8.61111111111111,
      "grad_norm": 3.767841100692749,
      "learning_rate": 3.16358024691358e-05,
      "loss": 0.2071,
      "step": 775
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 3.087602138519287,
      "learning_rate": 3.08641975308642e-05,
      "loss": 0.212,
      "step": 800
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.7687747035573123,
      "eval_loss": 0.6767174005508423,
      "eval_runtime": 31.6137,
      "eval_samples_per_second": 32.011,
      "eval_steps_per_second": 0.506,
      "step": 810
    },
    {
      "epoch": 9.166666666666666,
      "grad_norm": 2.6417057514190674,
      "learning_rate": 3.0092592592592593e-05,
      "loss": 0.1927,
      "step": 825
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 3.398786783218384,
      "learning_rate": 2.9320987654320992e-05,
      "loss": 0.1881,
      "step": 850
    },
    {
      "epoch": 9.722222222222221,
      "grad_norm": 2.6754283905029297,
      "learning_rate": 2.8549382716049384e-05,
      "loss": 0.1878,
      "step": 875
    },
    {
      "epoch": 10.0,
      "grad_norm": 4.713382720947266,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.1972,
      "step": 900
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.7569169960474308,
      "eval_loss": 0.6680033206939697,
      "eval_runtime": 31.73,
      "eval_samples_per_second": 31.894,
      "eval_steps_per_second": 0.504,
      "step": 900
    },
    {
      "epoch": 10.277777777777779,
      "grad_norm": 1.9474177360534668,
      "learning_rate": 2.700617283950617e-05,
      "loss": 0.1762,
      "step": 925
    },
    {
      "epoch": 10.555555555555555,
      "grad_norm": 3.6200313568115234,
      "learning_rate": 2.623456790123457e-05,
      "loss": 0.1768,
      "step": 950
    },
    {
      "epoch": 10.833333333333334,
      "grad_norm": 1.886972188949585,
      "learning_rate": 2.5462962962962965e-05,
      "loss": 0.174,
      "step": 975
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.775691699604743,
      "eval_loss": 0.6353378891944885,
      "eval_runtime": 31.1908,
      "eval_samples_per_second": 32.445,
      "eval_steps_per_second": 0.513,
      "step": 990
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 2.2933850288391113,
      "learning_rate": 2.4691358024691357e-05,
      "loss": 0.1651,
      "step": 1000
    },
    {
      "epoch": 11.38888888888889,
      "grad_norm": 1.7583023309707642,
      "learning_rate": 2.3919753086419755e-05,
      "loss": 0.163,
      "step": 1025
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 1.8716415166854858,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.1687,
      "step": 1050
    },
    {
      "epoch": 11.944444444444445,
      "grad_norm": 1.6115872859954834,
      "learning_rate": 2.2376543209876542e-05,
      "loss": 0.1637,
      "step": 1075
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.775691699604743,
      "eval_loss": 0.638940155506134,
      "eval_runtime": 30.9951,
      "eval_samples_per_second": 32.65,
      "eval_steps_per_second": 0.516,
      "step": 1080
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 2.395264148712158,
      "learning_rate": 2.1604938271604937e-05,
      "loss": 0.1578,
      "step": 1100
    },
    {
      "epoch": 12.5,
      "grad_norm": 3.230558156967163,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.1634,
      "step": 1125
    },
    {
      "epoch": 12.777777777777779,
      "grad_norm": 1.9428215026855469,
      "learning_rate": 2.006172839506173e-05,
      "loss": 0.1558,
      "step": 1150
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.7885375494071146,
      "eval_loss": 0.6259772777557373,
      "eval_runtime": 30.7544,
      "eval_samples_per_second": 32.906,
      "eval_steps_per_second": 0.52,
      "step": 1170
    },
    {
      "epoch": 13.055555555555555,
      "grad_norm": 2.128429889678955,
      "learning_rate": 1.9290123456790123e-05,
      "loss": 0.1535,
      "step": 1175
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 2.604618549346924,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.1594,
      "step": 1200
    },
    {
      "epoch": 13.61111111111111,
      "grad_norm": 1.9983032941818237,
      "learning_rate": 1.7746913580246917e-05,
      "loss": 0.1532,
      "step": 1225
    },
    {
      "epoch": 13.88888888888889,
      "grad_norm": 1.9643396139144897,
      "learning_rate": 1.697530864197531e-05,
      "loss": 0.153,
      "step": 1250
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.7796442687747036,
      "eval_loss": 0.6343194842338562,
      "eval_runtime": 31.603,
      "eval_samples_per_second": 32.022,
      "eval_steps_per_second": 0.506,
      "step": 1260
    },
    {
      "epoch": 14.166666666666666,
      "grad_norm": 1.8285529613494873,
      "learning_rate": 1.6203703703703704e-05,
      "loss": 0.1558,
      "step": 1275
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 2.246311664581299,
      "learning_rate": 1.54320987654321e-05,
      "loss": 0.1487,
      "step": 1300
    },
    {
      "epoch": 14.722222222222221,
      "grad_norm": 3.1947481632232666,
      "learning_rate": 1.4660493827160496e-05,
      "loss": 0.1444,
      "step": 1325
    },
    {
      "epoch": 15.0,
      "grad_norm": 2.330897331237793,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.1514,
      "step": 1350
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.7766798418972332,
      "eval_loss": 0.6341674327850342,
      "eval_runtime": 31.4549,
      "eval_samples_per_second": 32.173,
      "eval_steps_per_second": 0.509,
      "step": 1350
    },
    {
      "epoch": 15.277777777777779,
      "grad_norm": 1.669419765472412,
      "learning_rate": 1.3117283950617285e-05,
      "loss": 0.1475,
      "step": 1375
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 1.8953179121017456,
      "learning_rate": 1.2345679012345678e-05,
      "loss": 0.1416,
      "step": 1400
    },
    {
      "epoch": 15.833333333333334,
      "grad_norm": 1.6065765619277954,
      "learning_rate": 1.1574074074074075e-05,
      "loss": 0.1419,
      "step": 1425
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.7747035573122529,
      "eval_loss": 0.6450176239013672,
      "eval_runtime": 31.9216,
      "eval_samples_per_second": 31.703,
      "eval_steps_per_second": 0.501,
      "step": 1440
    },
    {
      "epoch": 16.11111111111111,
      "grad_norm": 2.2864298820495605,
      "learning_rate": 1.0802469135802469e-05,
      "loss": 0.1403,
      "step": 1450
    },
    {
      "epoch": 16.38888888888889,
      "grad_norm": 2.1860485076904297,
      "learning_rate": 1.0030864197530866e-05,
      "loss": 0.1395,
      "step": 1475
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 1.2659510374069214,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.1395,
      "step": 1500
    },
    {
      "epoch": 16.944444444444443,
      "grad_norm": 1.1790435314178467,
      "learning_rate": 8.487654320987654e-06,
      "loss": 0.1411,
      "step": 1525
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.782608695652174,
      "eval_loss": 0.6400145292282104,
      "eval_runtime": 32.2874,
      "eval_samples_per_second": 31.343,
      "eval_steps_per_second": 0.496,
      "step": 1530
    },
    {
      "epoch": 17.22222222222222,
      "grad_norm": 1.525313377380371,
      "learning_rate": 7.71604938271605e-06,
      "loss": 0.1419,
      "step": 1550
    },
    {
      "epoch": 17.5,
      "grad_norm": 1.0568948984146118,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.1328,
      "step": 1575
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 1.9466252326965332,
      "learning_rate": 6.172839506172839e-06,
      "loss": 0.1303,
      "step": 1600
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.7786561264822134,
      "eval_loss": 0.6327863335609436,
      "eval_runtime": 31.5918,
      "eval_samples_per_second": 32.034,
      "eval_steps_per_second": 0.506,
      "step": 1620
    },
    {
      "epoch": 18.055555555555557,
      "grad_norm": 1.1328738927841187,
      "learning_rate": 5.401234567901234e-06,
      "loss": 0.1362,
      "step": 1625
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 1.676224708557129,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.1378,
      "step": 1650
    },
    {
      "epoch": 18.61111111111111,
      "grad_norm": 1.8911770582199097,
      "learning_rate": 3.858024691358025e-06,
      "loss": 0.1329,
      "step": 1675
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 1.0110254287719727,
      "learning_rate": 3.0864197530864196e-06,
      "loss": 0.1316,
      "step": 1700
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.7865612648221344,
      "eval_loss": 0.6286635994911194,
      "eval_runtime": 30.9151,
      "eval_samples_per_second": 32.735,
      "eval_steps_per_second": 0.518,
      "step": 1710
    },
    {
      "epoch": 19.166666666666668,
      "grad_norm": 1.2502254247665405,
      "learning_rate": 2.3148148148148148e-06,
      "loss": 0.1287,
      "step": 1725
    },
    {
      "epoch": 19.444444444444443,
      "grad_norm": 1.2474185228347778,
      "learning_rate": 1.5432098765432098e-06,
      "loss": 0.1332,
      "step": 1750
    },
    {
      "epoch": 19.72222222222222,
      "grad_norm": 1.0085457563400269,
      "learning_rate": 7.716049382716049e-07,
      "loss": 0.1305,
      "step": 1775
    },
    {
      "epoch": 20.0,
      "grad_norm": 1.9045767784118652,
      "learning_rate": 0.0,
      "loss": 0.1303,
      "step": 1800
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.7835968379446641,
      "eval_loss": 0.6264516711235046,
      "eval_runtime": 31.0405,
      "eval_samples_per_second": 32.603,
      "eval_steps_per_second": 0.515,
      "step": 1800
    },
    {
      "epoch": 20.0,
      "step": 1800,
      "total_flos": 0.0,
      "train_loss": 0.3401869531472524,
      "train_runtime": 10681.4553,
      "train_samples_per_second": 10.729,
      "train_steps_per_second": 0.169
    }
  ],
  "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
}