{
  "best_metric": 0.9210770013650191,
  "best_model_checkpoint": "/home/hoan/projects/nsfw/models/efficientvit_l1.r224_in1k/run-fx97n43d/checkpoint-488",
  "epoch": 22.96470588235294,
  "eval_steps": 500,
  "global_step": 488,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.61,
      "learning_rate": 1.5510039521887183e-05,
      "loss": 1.478,
      "step": 13
    },
    {
      "epoch": 0.99,
      "eval_accuracy": 0.8287139689578714,
      "eval_f1": 0.7513704434640557,
      "eval_loss": 1.0341465473175049,
      "eval_precision": 0.7699235349840203,
      "eval_recall": 0.8287139689578714,
      "eval_runtime": 8.3797,
      "eval_samples_per_second": 430.565,
      "eval_steps_per_second": 3.461,
      "step": 21
    },
    {
      "epoch": 1.22,
      "learning_rate": 3.1020079043774365e-05,
      "loss": 1.1162,
      "step": 26
    },
    {
      "epoch": 1.84,
      "learning_rate": 4.6530118565661544e-05,
      "loss": 0.9424,
      "step": 39
    },
    {
      "epoch": 1.98,
      "eval_accuracy": 0.8705654101995566,
      "eval_f1": 0.8389699621462066,
      "eval_loss": 0.9252289533615112,
      "eval_precision": 0.8517874429530109,
      "eval_recall": 0.8705654101995566,
      "eval_runtime": 7.8428,
      "eval_samples_per_second": 460.041,
      "eval_steps_per_second": 3.698,
      "step": 42
    },
    {
      "epoch": 2.45,
      "learning_rate": 6.204015808754873e-05,
      "loss": 0.8696,
      "step": 52
    },
    {
      "epoch": 2.96,
      "eval_accuracy": 0.9068736141906873,
      "eval_f1": 0.8946612312334215,
      "eval_loss": 0.9877755045890808,
      "eval_precision": 0.8961631624753195,
      "eval_recall": 0.9068736141906873,
      "eval_runtime": 7.579,
      "eval_samples_per_second": 476.05,
      "eval_steps_per_second": 3.826,
      "step": 63
    },
    {
      "epoch": 3.06,
      "learning_rate": 7.755019760943591e-05,
      "loss": 0.8273,
      "step": 65
    },
    {
      "epoch": 3.67,
      "learning_rate": 9.306023713132309e-05,
      "loss": 0.7988,
      "step": 78
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9135254988913526,
      "eval_f1": 0.9080269220857267,
      "eval_loss": 0.8059695363044739,
      "eval_precision": 0.9037754375795349,
      "eval_recall": 0.9135254988913526,
      "eval_runtime": 7.9757,
      "eval_samples_per_second": 452.374,
      "eval_steps_per_second": 3.636,
      "step": 85
    },
    {
      "epoch": 4.28,
      "learning_rate": 0.00010857027665321027,
      "loss": 0.798,
      "step": 91
    },
    {
      "epoch": 4.89,
      "learning_rate": 0.00011326347231548932,
      "loss": 0.774,
      "step": 104
    },
    {
      "epoch": 4.99,
      "eval_accuracy": 0.9118625277161863,
      "eval_f1": 0.9080705370958632,
      "eval_loss": 0.8073797225952148,
      "eval_precision": 0.9073476317414986,
      "eval_recall": 0.9118625277161863,
      "eval_runtime": 8.2483,
      "eval_samples_per_second": 437.426,
      "eval_steps_per_second": 3.516,
      "step": 106
    },
    {
      "epoch": 5.51,
      "learning_rate": 0.00011287035231995169,
      "loss": 0.7733,
      "step": 117
    },
    {
      "epoch": 5.98,
      "eval_accuracy": 0.9154656319290465,
      "eval_f1": 0.9128434775291835,
      "eval_loss": 0.8052764534950256,
      "eval_precision": 0.910962080004432,
      "eval_recall": 0.9154656319290465,
      "eval_runtime": 8.2107,
      "eval_samples_per_second": 439.429,
      "eval_steps_per_second": 3.532,
      "step": 127
    },
    {
      "epoch": 6.12,
      "learning_rate": 0.00011214989390672101,
      "loss": 0.7647,
      "step": 130
    },
    {
      "epoch": 6.73,
      "learning_rate": 0.0001111062934759132,
      "loss": 0.7583,
      "step": 143
    },
    {
      "epoch": 6.96,
      "eval_accuracy": 0.9182372505543237,
      "eval_f1": 0.9138869052547461,
      "eval_loss": 0.7971659898757935,
      "eval_precision": 0.9104077481882505,
      "eval_recall": 0.9182372505543237,
      "eval_runtime": 7.619,
      "eval_samples_per_second": 473.555,
      "eval_steps_per_second": 3.806,
      "step": 148
    },
    {
      "epoch": 7.34,
      "learning_rate": 0.00010974562960873211,
      "loss": 0.7547,
      "step": 156
    },
    {
      "epoch": 7.95,
      "learning_rate": 0.00010807582766201317,
      "loss": 0.7439,
      "step": 169
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9154656319290465,
      "eval_f1": 0.9108299991634333,
      "eval_loss": 0.8041123151779175,
      "eval_precision": 0.907292999433662,
      "eval_recall": 0.9154656319290465,
      "eval_runtime": 7.8549,
      "eval_samples_per_second": 459.332,
      "eval_steps_per_second": 3.692,
      "step": 170
    },
    {
      "epoch": 8.56,
      "learning_rate": 0.00010610661360599156,
      "loss": 0.738,
      "step": 182
    },
    {
      "epoch": 8.99,
      "eval_accuracy": 0.917960088691796,
      "eval_f1": 0.9113645109824469,
      "eval_loss": 0.7966068387031555,
      "eval_precision": 0.9076893913635398,
      "eval_recall": 0.917960088691796,
      "eval_runtime": 7.6442,
      "eval_samples_per_second": 471.993,
      "eval_steps_per_second": 3.794,
      "step": 191
    },
    {
      "epoch": 9.18,
      "learning_rate": 0.00010384945737417238,
      "loss": 0.7397,
      "step": 195
    },
    {
      "epoch": 9.79,
      "learning_rate": 0.00010131750605526875,
      "loss": 0.735,
      "step": 208
    },
    {
      "epoch": 9.98,
      "eval_accuracy": 0.9201773835920177,
      "eval_f1": 0.9121076710952039,
      "eval_loss": 0.804332435131073,
      "eval_precision": 0.9097350024636609,
      "eval_recall": 0.9201773835920177,
      "eval_runtime": 8.136,
      "eval_samples_per_second": 443.464,
      "eval_steps_per_second": 3.564,
      "step": 212
    },
    {
      "epoch": 10.4,
      "learning_rate": 9.85255073163396e-05,
      "loss": 0.7292,
      "step": 221
    },
    {
      "epoch": 10.96,
      "eval_accuracy": 0.8991130820399114,
      "eval_f1": 0.901414211525136,
      "eval_loss": 0.8244433999061584,
      "eval_precision": 0.906232190945862,
      "eval_recall": 0.8991130820399114,
      "eval_runtime": 8.1006,
      "eval_samples_per_second": 445.397,
      "eval_steps_per_second": 3.58,
      "step": 233
    },
    {
      "epoch": 11.01,
      "learning_rate": 9.548972350315878e-05,
      "loss": 0.7325,
      "step": 234
    },
    {
      "epoch": 11.62,
      "learning_rate": 9.222783691814877e-05,
      "loss": 0.7236,
      "step": 247
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.893569844789357,
      "eval_f1": 0.8989946911162718,
      "eval_loss": 0.8460181951522827,
      "eval_precision": 0.9100742788902317,
      "eval_recall": 0.893569844789357,
      "eval_runtime": 7.883,
      "eval_samples_per_second": 457.695,
      "eval_steps_per_second": 3.679,
      "step": 255
    },
    {
      "epoch": 12.24,
      "learning_rate": 8.875884682759899e-05,
      "loss": 0.7165,
      "step": 260
    },
    {
      "epoch": 12.85,
      "learning_rate": 8.510295879806365e-05,
      "loss": 0.7154,
      "step": 273
    },
    {
      "epoch": 12.99,
      "eval_accuracy": 0.9121396895787139,
      "eval_f1": 0.9074187854075062,
      "eval_loss": 0.8101195096969604,
      "eval_precision": 0.9037000363072012,
      "eval_recall": 0.9121396895787139,
      "eval_runtime": 7.9381,
      "eval_samples_per_second": 454.518,
      "eval_steps_per_second": 3.653,
      "step": 276
    },
    {
      "epoch": 13.46,
      "learning_rate": 8.12814670065128e-05,
      "loss": 0.7157,
      "step": 286
    },
    {
      "epoch": 13.98,
      "eval_accuracy": 0.914079822616408,
      "eval_f1": 0.9064691154662861,
      "eval_loss": 0.8145021200180054,
      "eval_precision": 0.9073091371333267,
      "eval_recall": 0.914079822616408,
      "eval_runtime": 7.9076,
      "eval_samples_per_second": 456.272,
      "eval_steps_per_second": 3.667,
      "step": 297
    },
    {
      "epoch": 14.07,
      "learning_rate": 7.731663020973596e-05,
      "loss": 0.7141,
      "step": 299
    },
    {
      "epoch": 14.68,
      "learning_rate": 7.323154209542975e-05,
      "loss": 0.713,
      "step": 312
    },
    {
      "epoch": 14.96,
      "eval_accuracy": 0.914079822616408,
      "eval_f1": 0.904905291376415,
      "eval_loss": 0.8159310817718506,
      "eval_precision": 0.9062799076312693,
      "eval_recall": 0.914079822616408,
      "eval_runtime": 7.2278,
      "eval_samples_per_second": 499.183,
      "eval_steps_per_second": 4.012,
      "step": 318
    },
    {
      "epoch": 15.29,
      "learning_rate": 6.904999677012627e-05,
      "loss": 0.7102,
      "step": 325
    },
    {
      "epoch": 15.91,
      "learning_rate": 6.479635016744447e-05,
      "loss": 0.7138,
      "step": 338
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.9132483370288248,
      "eval_f1": 0.9123090321473447,
      "eval_loss": 0.8077041506767273,
      "eval_precision": 0.9126418966766847,
      "eval_recall": 0.9132483370288248,
      "eval_runtime": 7.5812,
      "eval_samples_per_second": 475.917,
      "eval_steps_per_second": 3.825,
      "step": 340
    },
    {
      "epoch": 16.52,
      "learning_rate": 6.0495378183908996e-05,
      "loss": 0.7031,
      "step": 351
    },
    {
      "epoch": 16.99,
      "eval_accuracy": 0.9210088691796009,
      "eval_f1": 0.9172330986693904,
      "eval_loss": 0.8018594980239868,
      "eval_precision": 0.914982087273544,
      "eval_recall": 0.9210088691796009,
      "eval_runtime": 7.7209,
      "eval_samples_per_second": 467.301,
      "eval_steps_per_second": 3.756,
      "step": 361
    },
    {
      "epoch": 17.13,
      "learning_rate": 5.617213236864303e-05,
      "loss": 0.7066,
      "step": 364
    },
    {
      "epoch": 17.74,
      "learning_rate": 5.185179400748782e-05,
      "loss": 0.7025,
      "step": 377
    },
    {
      "epoch": 17.98,
      "eval_accuracy": 0.9157427937915743,
      "eval_f1": 0.9121013658355167,
      "eval_loss": 0.8063293099403381,
      "eval_precision": 0.9101548570689432,
      "eval_recall": 0.9157427937915743,
      "eval_runtime": 7.6756,
      "eval_samples_per_second": 470.059,
      "eval_steps_per_second": 3.778,
      "step": 382
    },
    {
      "epoch": 18.35,
      "learning_rate": 4.7559527451454613e-05,
      "loss": 0.7023,
      "step": 390
    },
    {
      "epoch": 18.96,
      "learning_rate": 4.332033354381553e-05,
      "loss": 0.7039,
      "step": 403
    },
    {
      "epoch": 18.96,
      "eval_accuracy": 0.9165742793791575,
      "eval_f1": 0.9110959153728131,
      "eval_loss": 0.8079176545143127,
      "eval_precision": 0.9076920456818248,
      "eval_recall": 0.9165742793791575,
      "eval_runtime": 7.7227,
      "eval_samples_per_second": 467.196,
      "eval_steps_per_second": 3.755,
      "step": 403
    },
    {
      "epoch": 19.58,
      "learning_rate": 3.915890399956582e-05,
      "loss": 0.7007,
      "step": 416
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8999445676274944,
      "eval_f1": 0.9018109542637969,
      "eval_loss": 0.8286749720573425,
      "eval_precision": 0.9064557605321089,
      "eval_recall": 0.8999445676274944,
      "eval_runtime": 7.6911,
      "eval_samples_per_second": 469.114,
      "eval_steps_per_second": 3.771,
      "step": 425
    },
    {
      "epoch": 20.19,
      "learning_rate": 3.509947758544222e-05,
      "loss": 0.7008,
      "step": 429
    },
    {
      "epoch": 20.8,
      "learning_rate": 3.1165698938195404e-05,
      "loss": 0.7013,
      "step": 442
    },
    {
      "epoch": 20.99,
      "eval_accuracy": 0.9196230598669624,
      "eval_f1": 0.9160103903036274,
      "eval_loss": 0.8033446669578552,
      "eval_precision": 0.9157986272829293,
      "eval_recall": 0.9196230598669624,
      "eval_runtime": 7.6705,
      "eval_samples_per_second": 470.371,
      "eval_steps_per_second": 3.781,
      "step": 446
    },
    {
      "epoch": 21.41,
      "learning_rate": 2.738048084344645e-05,
      "loss": 0.698,
      "step": 455
    },
    {
      "epoch": 21.98,
      "eval_accuracy": 0.9149113082039911,
      "eval_f1": 0.9105505264344369,
      "eval_loss": 0.8064530491828918,
      "eval_precision": 0.9079365511507042,
      "eval_recall": 0.9149113082039911,
      "eval_runtime": 7.8595,
      "eval_samples_per_second": 459.061,
      "eval_steps_per_second": 3.69,
      "step": 467
    },
    {
      "epoch": 22.02,
      "learning_rate": 2.3765870777302054e-05,
      "loss": 0.6984,
      "step": 468
    },
    {
      "epoch": 22.64,
      "learning_rate": 2.0342922488073023e-05,
      "loss": 0.6962,
      "step": 481
    },
    {
      "epoch": 22.96,
      "eval_accuracy": 0.9254434589800443,
      "eval_f1": 0.9210770013650191,
      "eval_loss": 0.7944092154502869,
      "eval_precision": 0.9202816675752099,
      "eval_recall": 0.9254434589800443,
      "eval_runtime": 7.8528,
      "eval_samples_per_second": 459.456,
      "eval_steps_per_second": 3.693,
      "step": 488
    }
  ],
  "logging_steps": 13,
  "max_steps": 630,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 30,
  "save_steps": 500,
  "total_flos": 0.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": {
    "_wandb": {},
    "assignments": {},
    "dropout_rate": 0.18565388367817143,
    "learning_rate": 0.00011334259650609864,
    "metric": "eval/loss",
    "use_mish": true
  }
}