{
  "best_metric": 0.1197938472032547,
  "best_model_checkpoint": "./vit-base-brain-dementia-detection1/checkpoint-2500",
  "epoch": 15.625,
  "eval_steps": 500,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1953125,
      "grad_norm": 2.9218900203704834,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.3496,
      "step": 50
    },
    {
      "epoch": 0.390625,
      "grad_norm": 2.503772258758545,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.1755,
      "step": 100
    },
    {
      "epoch": 0.5859375,
      "grad_norm": 2.9283928871154785,
      "learning_rate": 6e-06,
      "loss": 1.0408,
      "step": 150
    },
    {
      "epoch": 0.78125,
      "grad_norm": 2.1896207332611084,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.9946,
      "step": 200
    },
    {
      "epoch": 0.9765625,
      "grad_norm": 2.5399117469787598,
      "learning_rate": 1e-05,
      "loss": 0.904,
      "step": 250
    },
    {
      "epoch": 1.171875,
      "grad_norm": 2.1824874877929688,
      "learning_rate": 1.2e-05,
      "loss": 0.8886,
      "step": 300
    },
    {
      "epoch": 1.3671875,
      "grad_norm": 3.012261390686035,
      "learning_rate": 1.4e-05,
      "loss": 0.8417,
      "step": 350
    },
    {
      "epoch": 1.5625,
      "grad_norm": 2.7216336727142334,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.7733,
      "step": 400
    },
    {
      "epoch": 1.7578125,
      "grad_norm": 9.597420692443848,
      "learning_rate": 1.8e-05,
      "loss": 0.7711,
      "step": 450
    },
    {
      "epoch": 1.953125,
      "grad_norm": 3.229459524154663,
      "learning_rate": 2e-05,
      "loss": 0.7193,
      "step": 500
    },
    {
      "epoch": 1.953125,
      "eval_accuracy": 0.6591796875,
      "eval_loss": 0.7949748039245605,
      "eval_runtime": 6.2299,
      "eval_samples_per_second": 164.369,
      "eval_steps_per_second": 20.546,
      "step": 500
    },
    {
      "epoch": 2.1484375,
      "grad_norm": 13.59356689453125,
      "learning_rate": 1.9783549783549785e-05,
      "loss": 0.6266,
      "step": 550
    },
    {
      "epoch": 2.34375,
      "grad_norm": 10.799866676330566,
      "learning_rate": 1.9567099567099568e-05,
      "loss": 0.6088,
      "step": 600
    },
    {
      "epoch": 2.5390625,
      "grad_norm": 14.872405052185059,
      "learning_rate": 1.9350649350649354e-05,
      "loss": 0.5726,
      "step": 650
    },
    {
      "epoch": 2.734375,
      "grad_norm": 13.640074729919434,
      "learning_rate": 1.9134199134199138e-05,
      "loss": 0.529,
      "step": 700
    },
    {
      "epoch": 2.9296875,
      "grad_norm": 15.788460731506348,
      "learning_rate": 1.891774891774892e-05,
      "loss": 0.4271,
      "step": 750
    },
    {
      "epoch": 3.125,
      "grad_norm": 4.504462242126465,
      "learning_rate": 1.8701298701298704e-05,
      "loss": 0.3618,
      "step": 800
    },
    {
      "epoch": 3.3203125,
      "grad_norm": 3.5319948196411133,
      "learning_rate": 1.8484848484848487e-05,
      "loss": 0.3656,
      "step": 850
    },
    {
      "epoch": 3.515625,
      "grad_norm": 0.7339128851890564,
      "learning_rate": 1.826839826839827e-05,
      "loss": 0.2925,
      "step": 900
    },
    {
      "epoch": 3.7109375,
      "grad_norm": 13.2314453125,
      "learning_rate": 1.8051948051948053e-05,
      "loss": 0.2769,
      "step": 950
    },
    {
      "epoch": 3.90625,
      "grad_norm": 13.87197208404541,
      "learning_rate": 1.7835497835497836e-05,
      "loss": 0.249,
      "step": 1000
    },
    {
      "epoch": 3.90625,
      "eval_accuracy": 0.90234375,
      "eval_loss": 0.3422887325286865,
      "eval_runtime": 5.8006,
      "eval_samples_per_second": 176.535,
      "eval_steps_per_second": 22.067,
      "step": 1000
    },
    {
      "epoch": 4.1015625,
      "grad_norm": 10.137914657592773,
      "learning_rate": 1.761904761904762e-05,
      "loss": 0.2877,
      "step": 1050
    },
    {
      "epoch": 4.296875,
      "grad_norm": 20.356388092041016,
      "learning_rate": 1.7402597402597403e-05,
      "loss": 0.2081,
      "step": 1100
    },
    {
      "epoch": 4.4921875,
      "grad_norm": 15.127123832702637,
      "learning_rate": 1.718614718614719e-05,
      "loss": 0.1865,
      "step": 1150
    },
    {
      "epoch": 4.6875,
      "grad_norm": 0.5982966423034668,
      "learning_rate": 1.6969696969696972e-05,
      "loss": 0.1477,
      "step": 1200
    },
    {
      "epoch": 4.8828125,
      "grad_norm": 9.74530029296875,
      "learning_rate": 1.6753246753246756e-05,
      "loss": 0.1313,
      "step": 1250
    },
    {
      "epoch": 5.078125,
      "grad_norm": 0.6074052453041077,
      "learning_rate": 1.653679653679654e-05,
      "loss": 0.1095,
      "step": 1300
    },
    {
      "epoch": 5.2734375,
      "grad_norm": 0.18593043088912964,
      "learning_rate": 1.6320346320346322e-05,
      "loss": 0.1199,
      "step": 1350
    },
    {
      "epoch": 5.46875,
      "grad_norm": 0.22421252727508545,
      "learning_rate": 1.6103896103896105e-05,
      "loss": 0.1042,
      "step": 1400
    },
    {
      "epoch": 5.6640625,
      "grad_norm": 1.1787017583847046,
      "learning_rate": 1.5887445887445888e-05,
      "loss": 0.072,
      "step": 1450
    },
    {
      "epoch": 5.859375,
      "grad_norm": 22.340038299560547,
      "learning_rate": 1.567099567099567e-05,
      "loss": 0.0774,
      "step": 1500
    },
    {
      "epoch": 5.859375,
      "eval_accuracy": 0.94921875,
      "eval_loss": 0.18452830612659454,
      "eval_runtime": 5.6092,
      "eval_samples_per_second": 182.558,
      "eval_steps_per_second": 22.82,
      "step": 1500
    },
    {
      "epoch": 6.0546875,
      "grad_norm": 22.49978256225586,
      "learning_rate": 1.5454545454545454e-05,
      "loss": 0.0376,
      "step": 1550
    },
    {
      "epoch": 6.25,
      "grad_norm": 1.0888569355010986,
      "learning_rate": 1.523809523809524e-05,
      "loss": 0.0541,
      "step": 1600
    },
    {
      "epoch": 6.4453125,
      "grad_norm": 0.13725623488426208,
      "learning_rate": 1.5021645021645024e-05,
      "loss": 0.0516,
      "step": 1650
    },
    {
      "epoch": 6.640625,
      "grad_norm": 0.20473556220531464,
      "learning_rate": 1.4805194805194807e-05,
      "loss": 0.0326,
      "step": 1700
    },
    {
      "epoch": 6.8359375,
      "grad_norm": 0.25789499282836914,
      "learning_rate": 1.458874458874459e-05,
      "loss": 0.0554,
      "step": 1750
    },
    {
      "epoch": 7.03125,
      "grad_norm": 0.08650106191635132,
      "learning_rate": 1.4372294372294374e-05,
      "loss": 0.0259,
      "step": 1800
    },
    {
      "epoch": 7.2265625,
      "grad_norm": 0.0898546352982521,
      "learning_rate": 1.4155844155844157e-05,
      "loss": 0.0256,
      "step": 1850
    },
    {
      "epoch": 7.421875,
      "grad_norm": 0.07839500159025192,
      "learning_rate": 1.3939393939393942e-05,
      "loss": 0.0288,
      "step": 1900
    },
    {
      "epoch": 7.6171875,
      "grad_norm": 0.07453235238790512,
      "learning_rate": 1.3722943722943725e-05,
      "loss": 0.0178,
      "step": 1950
    },
    {
      "epoch": 7.8125,
      "grad_norm": 0.075274258852005,
      "learning_rate": 1.3506493506493508e-05,
      "loss": 0.0306,
      "step": 2000
    },
    {
      "epoch": 7.8125,
      "eval_accuracy": 0.95703125,
      "eval_loss": 0.18091866374015808,
      "eval_runtime": 5.5217,
      "eval_samples_per_second": 185.451,
      "eval_steps_per_second": 23.181,
      "step": 2000
    },
    {
      "epoch": 8.0078125,
      "grad_norm": 0.06450609862804413,
      "learning_rate": 1.3290043290043291e-05,
      "loss": 0.0161,
      "step": 2050
    },
    {
      "epoch": 8.203125,
      "grad_norm": 0.12140965461730957,
      "learning_rate": 1.3073593073593074e-05,
      "loss": 0.0223,
      "step": 2100
    },
    {
      "epoch": 8.3984375,
      "grad_norm": 0.05588414520025253,
      "learning_rate": 1.2857142857142859e-05,
      "loss": 0.0139,
      "step": 2150
    },
    {
      "epoch": 8.59375,
      "grad_norm": 0.054097168147563934,
      "learning_rate": 1.2640692640692642e-05,
      "loss": 0.0129,
      "step": 2200
    },
    {
      "epoch": 8.7890625,
      "grad_norm": 0.6485223770141602,
      "learning_rate": 1.2424242424242425e-05,
      "loss": 0.0193,
      "step": 2250
    },
    {
      "epoch": 8.984375,
      "grad_norm": 0.37905704975128174,
      "learning_rate": 1.2212121212121213e-05,
      "loss": 0.0265,
      "step": 2300
    },
    {
      "epoch": 9.1796875,
      "grad_norm": 0.05206324905157089,
      "learning_rate": 1.1995670995670996e-05,
      "loss": 0.0112,
      "step": 2350
    },
    {
      "epoch": 9.375,
      "grad_norm": 0.04438047483563423,
      "learning_rate": 1.1779220779220779e-05,
      "loss": 0.01,
      "step": 2400
    },
    {
      "epoch": 9.5703125,
      "grad_norm": 0.046205826103687286,
      "learning_rate": 1.1562770562770564e-05,
      "loss": 0.0102,
      "step": 2450
    },
    {
      "epoch": 9.765625,
      "grad_norm": 0.05957115814089775,
      "learning_rate": 1.1346320346320347e-05,
      "loss": 0.0099,
      "step": 2500
    },
    {
      "epoch": 9.765625,
      "eval_accuracy": 0.9716796875,
      "eval_loss": 0.1197938472032547,
      "eval_runtime": 5.5803,
      "eval_samples_per_second": 183.501,
      "eval_steps_per_second": 22.938,
      "step": 2500
    },
    {
      "epoch": 9.9609375,
      "grad_norm": 0.22725243866443634,
      "learning_rate": 1.112987012987013e-05,
      "loss": 0.0161,
      "step": 2550
    },
    {
      "epoch": 10.15625,
      "grad_norm": 0.040088605135679245,
      "learning_rate": 1.0913419913419913e-05,
      "loss": 0.009,
      "step": 2600
    },
    {
      "epoch": 10.3515625,
      "grad_norm": 0.0421510674059391,
      "learning_rate": 1.0696969696969696e-05,
      "loss": 0.0079,
      "step": 2650
    },
    {
      "epoch": 10.546875,
      "grad_norm": 0.04016215354204178,
      "learning_rate": 1.0480519480519481e-05,
      "loss": 0.008,
      "step": 2700
    },
    {
      "epoch": 10.7421875,
      "grad_norm": 0.03572896867990494,
      "learning_rate": 1.0264069264069264e-05,
      "loss": 0.0079,
      "step": 2750
    },
    {
      "epoch": 10.9375,
      "grad_norm": 0.033964116126298904,
      "learning_rate": 1.0047619047619048e-05,
      "loss": 0.0078,
      "step": 2800
    },
    {
      "epoch": 11.1328125,
      "grad_norm": 0.03913121297955513,
      "learning_rate": 9.831168831168832e-06,
      "loss": 0.0073,
      "step": 2850
    },
    {
      "epoch": 11.328125,
      "grad_norm": 0.03905245289206505,
      "learning_rate": 9.614718614718616e-06,
      "loss": 0.0067,
      "step": 2900
    },
    {
      "epoch": 11.5234375,
      "grad_norm": 0.03269728645682335,
      "learning_rate": 9.398268398268399e-06,
      "loss": 0.0068,
      "step": 2950
    },
    {
      "epoch": 11.71875,
      "grad_norm": 0.029141413047909737,
      "learning_rate": 9.181818181818184e-06,
      "loss": 0.0065,
      "step": 3000
    },
    {
      "epoch": 11.71875,
      "eval_accuracy": 0.96484375,
      "eval_loss": 0.1496964395046234,
      "eval_runtime": 6.3249,
      "eval_samples_per_second": 161.899,
      "eval_steps_per_second": 20.237,
      "step": 3000
    },
    {
      "epoch": 11.9140625,
      "grad_norm": 0.029785145074129105,
      "learning_rate": 8.965367965367967e-06,
      "loss": 0.0064,
      "step": 3050
    },
    {
      "epoch": 12.109375,
      "grad_norm": 0.028293365612626076,
      "learning_rate": 8.74891774891775e-06,
      "loss": 0.0059,
      "step": 3100
    },
    {
      "epoch": 12.3046875,
      "grad_norm": 0.026657408103346825,
      "learning_rate": 8.532467532467533e-06,
      "loss": 0.006,
      "step": 3150
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.028319496661424637,
      "learning_rate": 8.316017316017316e-06,
      "loss": 0.0057,
      "step": 3200
    },
    {
      "epoch": 12.6953125,
      "grad_norm": 0.025114014744758606,
      "learning_rate": 8.099567099567101e-06,
      "loss": 0.0054,
      "step": 3250
    },
    {
      "epoch": 12.890625,
      "grad_norm": 0.025485936552286148,
      "learning_rate": 7.883116883116884e-06,
      "loss": 0.0056,
      "step": 3300
    },
    {
      "epoch": 13.0859375,
      "grad_norm": 0.032143257558345795,
      "learning_rate": 7.666666666666667e-06,
      "loss": 0.0054,
      "step": 3350
    },
    {
      "epoch": 13.28125,
      "grad_norm": 0.02569337747991085,
      "learning_rate": 7.450216450216451e-06,
      "loss": 0.005,
      "step": 3400
    },
    {
      "epoch": 13.4765625,
      "grad_norm": 0.023684445768594742,
      "learning_rate": 7.233766233766234e-06,
      "loss": 0.005,
      "step": 3450
    },
    {
      "epoch": 13.671875,
      "grad_norm": 0.023895638063549995,
      "learning_rate": 7.017316017316018e-06,
      "loss": 0.0053,
      "step": 3500
    },
    {
      "epoch": 13.671875,
      "eval_accuracy": 0.966796875,
      "eval_loss": 0.1476988047361374,
      "eval_runtime": 5.5659,
      "eval_samples_per_second": 183.978,
      "eval_steps_per_second": 22.997,
      "step": 3500
    },
    {
      "epoch": 13.8671875,
      "grad_norm": 0.022996114566922188,
      "learning_rate": 6.8008658008658015e-06,
      "loss": 0.0048,
      "step": 3550
    },
    {
      "epoch": 14.0625,
      "grad_norm": 0.022488022223114967,
      "learning_rate": 6.5844155844155855e-06,
      "loss": 0.0048,
      "step": 3600
    },
    {
      "epoch": 14.2578125,
      "grad_norm": 0.038064874708652496,
      "learning_rate": 6.367965367965369e-06,
      "loss": 0.0045,
      "step": 3650
    },
    {
      "epoch": 14.453125,
      "grad_norm": 0.02063606306910515,
      "learning_rate": 6.151515151515152e-06,
      "loss": 0.0045,
      "step": 3700
    },
    {
      "epoch": 14.6484375,
      "grad_norm": 0.02058710902929306,
      "learning_rate": 5.935064935064936e-06,
      "loss": 0.0045,
      "step": 3750
    },
    {
      "epoch": 14.84375,
      "grad_norm": 0.019946841523051262,
      "learning_rate": 5.718614718614719e-06,
      "loss": 0.0045,
      "step": 3800
    },
    {
      "epoch": 15.0390625,
      "grad_norm": 0.02103234827518463,
      "learning_rate": 5.502164502164503e-06,
      "loss": 0.0042,
      "step": 3850
    },
    {
      "epoch": 15.234375,
      "grad_norm": 0.019999125972390175,
      "learning_rate": 5.285714285714286e-06,
      "loss": 0.0042,
      "step": 3900
    },
    {
      "epoch": 15.4296875,
      "grad_norm": 0.02068265713751316,
      "learning_rate": 5.069264069264069e-06,
      "loss": 0.0043,
      "step": 3950
    },
    {
      "epoch": 15.625,
      "grad_norm": 0.019038330763578415,
      "learning_rate": 4.852813852813853e-06,
      "loss": 0.004,
      "step": 4000
    },
    {
      "epoch": 15.625,
      "eval_accuracy": 0.962890625,
      "eval_loss": 0.1585080474615097,
      "eval_runtime": 5.7247,
      "eval_samples_per_second": 178.875,
      "eval_steps_per_second": 22.359,
      "step": 4000
    },
    {
      "epoch": 15.625,
      "step": 4000,
      "total_flos": 4.959576235966464e+18,
      "train_loss": 0.1988602912425995,
      "train_runtime": 917.7484,
      "train_samples_per_second": 89.262,
      "train_steps_per_second": 5.579
    }
  ],
  "logging_steps": 50,
  "max_steps": 5120,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.959576235966464e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}