{
  "best_metric": 1.696058988571167,
  "best_model_checkpoint": "detr-resnet50-leuk/checkpoint-4840",
  "epoch": 40.0,
  "global_step": 4840,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 5e-05,
      "loss": 3.1222,
      "step": 121
    },
    {
      "epoch": 1.0,
      "eval_loss": 2.3225085735321045,
      "eval_runtime": 19.3007,
      "eval_samples_per_second": 5.181,
      "eval_steps_per_second": 0.674,
      "step": 121
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0,
      "loss": 2.7011,
      "step": 242
    },
    {
      "epoch": 2.0,
      "eval_loss": 2.0469789505004883,
      "eval_runtime": 16.6883,
      "eval_samples_per_second": 5.992,
      "eval_steps_per_second": 0.779,
      "step": 242
    },
    {
      "epoch": 3.0,
      "learning_rate": 4e-05,
      "loss": 2.7217,
      "step": 363
    },
    {
      "epoch": 3.0,
      "eval_loss": 2.1208744049072266,
      "eval_runtime": 16.6967,
      "eval_samples_per_second": 5.989,
      "eval_steps_per_second": 0.779,
      "step": 363
    },
    {
      "epoch": 4.0,
      "learning_rate": 2e-05,
      "loss": 2.5204,
      "step": 484
    },
    {
      "epoch": 4.0,
      "eval_loss": 2.018857479095459,
      "eval_runtime": 16.6458,
      "eval_samples_per_second": 6.008,
      "eval_steps_per_second": 0.781,
      "step": 484
    },
    {
      "epoch": 5.0,
      "learning_rate": 0.0,
      "loss": 2.4531,
      "step": 605
    },
    {
      "epoch": 5.0,
      "eval_loss": 1.9292504787445068,
      "eval_runtime": 17.4457,
      "eval_samples_per_second": 5.732,
      "eval_steps_per_second": 0.745,
      "step": 605
    },
    {
      "epoch": 6.0,
      "learning_rate": 4e-05,
      "loss": 2.5448,
      "step": 726
    },
    {
      "epoch": 6.0,
      "eval_loss": 2.1573009490966797,
      "eval_runtime": 16.6325,
      "eval_samples_per_second": 6.012,
      "eval_steps_per_second": 0.782,
      "step": 726
    },
    {
      "epoch": 7.0,
      "learning_rate": 3e-05,
      "loss": 2.5266,
      "step": 847
    },
    {
      "epoch": 7.0,
      "eval_loss": 1.9386581182479858,
      "eval_runtime": 16.9375,
      "eval_samples_per_second": 5.904,
      "eval_steps_per_second": 0.768,
      "step": 847
    },
    {
      "epoch": 8.0,
      "learning_rate": 2e-05,
      "loss": 2.385,
      "step": 968
    },
    {
      "epoch": 8.0,
      "eval_loss": 1.9430466890335083,
      "eval_runtime": 16.7381,
      "eval_samples_per_second": 5.974,
      "eval_steps_per_second": 0.777,
      "step": 968
    },
    {
      "epoch": 9.0,
      "learning_rate": 1e-05,
      "loss": 2.3655,
      "step": 1089
    },
    {
      "epoch": 9.0,
      "eval_loss": 1.9069654941558838,
      "eval_runtime": 16.6356,
      "eval_samples_per_second": 6.011,
      "eval_steps_per_second": 0.781,
      "step": 1089
    },
    {
      "epoch": 10.0,
      "learning_rate": 0.0,
      "loss": 2.384,
      "step": 1210
    },
    {
      "epoch": 10.0,
      "eval_loss": 1.8772492408752441,
      "eval_runtime": 17.2204,
      "eval_samples_per_second": 5.807,
      "eval_steps_per_second": 0.755,
      "step": 1210
    },
    {
      "epoch": 11.0,
      "learning_rate": 4.5e-05,
      "loss": 2.4173,
      "step": 1331
    },
    {
      "epoch": 11.0,
      "eval_loss": 1.9407728910446167,
      "eval_runtime": 16.8832,
      "eval_samples_per_second": 5.923,
      "eval_steps_per_second": 0.77,
      "step": 1331
    },
    {
      "epoch": 12.0,
      "learning_rate": 4e-05,
      "loss": 2.4512,
      "step": 1452
    },
    {
      "epoch": 12.0,
      "eval_loss": 1.903834581375122,
      "eval_runtime": 16.6128,
      "eval_samples_per_second": 6.019,
      "eval_steps_per_second": 0.783,
      "step": 1452
    },
    {
      "epoch": 13.0,
      "learning_rate": 3.5e-05,
      "loss": 2.4599,
      "step": 1573
    },
    {
      "epoch": 13.0,
      "eval_loss": 2.049583673477173,
      "eval_runtime": 16.711,
      "eval_samples_per_second": 5.984,
      "eval_steps_per_second": 0.778,
      "step": 1573
    },
    {
      "epoch": 14.0,
      "learning_rate": 3e-05,
      "loss": 2.382,
      "step": 1694
    },
    {
      "epoch": 14.0,
      "eval_loss": 1.9043866395950317,
      "eval_runtime": 16.9051,
      "eval_samples_per_second": 5.915,
      "eval_steps_per_second": 0.769,
      "step": 1694
    },
    {
      "epoch": 15.0,
      "learning_rate": 2.5e-05,
      "loss": 2.3739,
      "step": 1815
    },
    {
      "epoch": 15.0,
      "eval_loss": 1.864935278892517,
      "eval_runtime": 17.2683,
      "eval_samples_per_second": 5.791,
      "eval_steps_per_second": 0.753,
      "step": 1815
    },
    {
      "epoch": 16.0,
      "learning_rate": 2e-05,
      "loss": 2.3066,
      "step": 1936
    },
    {
      "epoch": 16.0,
      "eval_loss": 1.8338868618011475,
      "eval_runtime": 16.6824,
      "eval_samples_per_second": 5.994,
      "eval_steps_per_second": 0.779,
      "step": 1936
    },
    {
      "epoch": 17.0,
      "learning_rate": 1.5e-05,
      "loss": 2.2597,
      "step": 2057
    },
    {
      "epoch": 17.0,
      "eval_loss": 1.7872685194015503,
      "eval_runtime": 16.6678,
      "eval_samples_per_second": 6.0,
      "eval_steps_per_second": 0.78,
      "step": 2057
    },
    {
      "epoch": 18.0,
      "learning_rate": 1e-05,
      "loss": 2.2254,
      "step": 2178
    },
    {
      "epoch": 18.0,
      "eval_loss": 1.8041014671325684,
      "eval_runtime": 17.0611,
      "eval_samples_per_second": 5.861,
      "eval_steps_per_second": 0.762,
      "step": 2178
    },
    {
      "epoch": 19.0,
      "learning_rate": 5e-06,
      "loss": 2.2674,
      "step": 2299
    },
    {
      "epoch": 19.0,
      "eval_loss": 1.794564962387085,
      "eval_runtime": 16.6625,
      "eval_samples_per_second": 6.002,
      "eval_steps_per_second": 0.78,
      "step": 2299
    },
    {
      "epoch": 20.0,
      "learning_rate": 0.0,
      "loss": 2.2109,
      "step": 2420
    },
    {
      "epoch": 20.0,
      "eval_loss": 1.7830548286437988,
      "eval_runtime": 16.6399,
      "eval_samples_per_second": 6.01,
      "eval_steps_per_second": 0.781,
      "step": 2420
    },
    {
      "epoch": 21.0,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 2.2812,
      "step": 2541
    },
    {
      "epoch": 21.0,
      "eval_loss": 1.8024096488952637,
      "eval_runtime": 17.0979,
      "eval_samples_per_second": 5.849,
      "eval_steps_per_second": 0.76,
      "step": 2541
    },
    {
      "epoch": 22.0,
      "learning_rate": 1.2e-05,
      "loss": 2.2648,
      "step": 2662
    },
    {
      "epoch": 22.0,
      "eval_loss": 1.7854045629501343,
      "eval_runtime": 17.5224,
      "eval_samples_per_second": 5.707,
      "eval_steps_per_second": 0.742,
      "step": 2662
    },
    {
      "epoch": 23.0,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.2161,
      "step": 2783
    },
    {
      "epoch": 23.0,
      "eval_loss": 1.7622228860855103,
      "eval_runtime": 17.3675,
      "eval_samples_per_second": 5.758,
      "eval_steps_per_second": 0.749,
      "step": 2783
    },
    {
      "epoch": 24.0,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.209,
      "step": 2904
    },
    {
      "epoch": 24.0,
      "eval_loss": 1.7543927431106567,
      "eval_runtime": 17.0326,
      "eval_samples_per_second": 5.871,
      "eval_steps_per_second": 0.763,
      "step": 2904
    },
    {
      "epoch": 25.0,
      "learning_rate": 0.0,
      "loss": 2.1905,
      "step": 3025
    },
    {
      "epoch": 25.0,
      "eval_loss": 1.7473363876342773,
      "eval_runtime": 17.5221,
      "eval_samples_per_second": 5.707,
      "eval_steps_per_second": 0.742,
      "step": 3025
    },
    {
      "epoch": 26.0,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 2.2166,
      "step": 3146
    },
    {
      "epoch": 26.0,
      "eval_loss": 1.7673826217651367,
      "eval_runtime": 16.6343,
      "eval_samples_per_second": 6.012,
      "eval_steps_per_second": 0.782,
      "step": 3146
    },
    {
      "epoch": 27.0,
      "learning_rate": 1e-05,
      "loss": 2.2108,
      "step": 3267
    },
    {
      "epoch": 27.0,
      "eval_loss": 1.7444889545440674,
      "eval_runtime": 16.7164,
      "eval_samples_per_second": 5.982,
      "eval_steps_per_second": 0.778,
      "step": 3267
    },
    {
      "epoch": 28.0,
      "learning_rate": 6.666666666666667e-06,
      "loss": 2.1813,
      "step": 3388
    },
    {
      "epoch": 28.0,
      "eval_loss": 1.732877254486084,
      "eval_runtime": 16.5798,
      "eval_samples_per_second": 6.031,
      "eval_steps_per_second": 0.784,
      "step": 3388
    },
    {
      "epoch": 29.0,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 2.1679,
      "step": 3509
    },
    {
      "epoch": 29.0,
      "eval_loss": 1.7286274433135986,
      "eval_runtime": 16.7429,
      "eval_samples_per_second": 5.973,
      "eval_steps_per_second": 0.776,
      "step": 3509
    },
    {
      "epoch": 30.0,
      "learning_rate": 0.0,
      "loss": 2.1481,
      "step": 3630
    },
    {
      "epoch": 30.0,
      "eval_loss": 1.7253737449645996,
      "eval_runtime": 16.881,
      "eval_samples_per_second": 5.924,
      "eval_steps_per_second": 0.77,
      "step": 3630
    },
    {
      "epoch": 31.0,
      "learning_rate": 1.1428571428571429e-05,
      "loss": 2.1713,
      "step": 3751
    },
    {
      "epoch": 31.0,
      "eval_loss": 1.7368007898330688,
      "eval_runtime": 17.8966,
      "eval_samples_per_second": 5.588,
      "eval_steps_per_second": 0.726,
      "step": 3751
    },
    {
      "epoch": 32.0,
      "learning_rate": 8.571428571428573e-06,
      "loss": 2.1471,
      "step": 3872
    },
    {
      "epoch": 32.0,
      "eval_loss": 1.7362391948699951,
      "eval_runtime": 16.9758,
      "eval_samples_per_second": 5.891,
      "eval_steps_per_second": 0.766,
      "step": 3872
    },
    {
      "epoch": 33.0,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 2.1537,
      "step": 3993
    },
    {
      "epoch": 33.0,
      "eval_loss": 1.7281136512756348,
      "eval_runtime": 17.1138,
      "eval_samples_per_second": 5.843,
      "eval_steps_per_second": 0.76,
      "step": 3993
    },
    {
      "epoch": 34.0,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 2.1347,
      "step": 4114
    },
    {
      "epoch": 34.0,
      "eval_loss": 1.720489501953125,
      "eval_runtime": 17.081,
      "eval_samples_per_second": 5.854,
      "eval_steps_per_second": 0.761,
      "step": 4114
    },
    {
      "epoch": 35.0,
      "learning_rate": 0.0,
      "loss": 2.129,
      "step": 4235
    },
    {
      "epoch": 35.0,
      "eval_loss": 1.710868239402771,
      "eval_runtime": 17.4345,
      "eval_samples_per_second": 5.736,
      "eval_steps_per_second": 0.746,
      "step": 4235
    },
    {
      "epoch": 36.0,
      "learning_rate": 1e-05,
      "loss": 2.1215,
      "step": 4356
    },
    {
      "epoch": 36.0,
      "eval_loss": 1.7227232456207275,
      "eval_runtime": 16.643,
      "eval_samples_per_second": 6.009,
      "eval_steps_per_second": 0.781,
      "step": 4356
    },
    {
      "epoch": 37.0,
      "learning_rate": 7.5e-06,
      "loss": 2.1425,
      "step": 4477
    },
    {
      "epoch": 37.0,
      "eval_loss": 1.7108997106552124,
      "eval_runtime": 16.8607,
      "eval_samples_per_second": 5.931,
      "eval_steps_per_second": 0.771,
      "step": 4477
    },
    {
      "epoch": 38.0,
      "learning_rate": 5e-06,
      "loss": 2.1106,
      "step": 4598
    },
    {
      "epoch": 38.0,
      "eval_loss": 1.6993489265441895,
      "eval_runtime": 16.7213,
      "eval_samples_per_second": 5.98,
      "eval_steps_per_second": 0.777,
      "step": 4598
    },
    {
      "epoch": 39.0,
      "learning_rate": 2.5e-06,
      "loss": 2.0987,
      "step": 4719
    },
    {
      "epoch": 39.0,
      "eval_loss": 1.6982134580612183,
      "eval_runtime": 17.0485,
      "eval_samples_per_second": 5.866,
      "eval_steps_per_second": 0.763,
      "step": 4719
    },
    {
      "epoch": 40.0,
      "learning_rate": 0.0,
      "loss": 2.1259,
      "step": 4840
    },
    {
      "epoch": 40.0,
      "eval_loss": 1.696058988571167,
      "eval_runtime": 17.3421,
      "eval_samples_per_second": 5.766,
      "eval_steps_per_second": 0.75,
      "step": 4840
    },
    {
      "epoch": 40.0,
      "step": 4840,
      "total_flos": 1.84657635477504e+19,
      "train_loss": 0.2649796856336357,
      "train_runtime": 1315.7115,
      "train_samples_per_second": 29.368,
      "train_steps_per_second": 3.679
    }
  ],
  "max_steps": 4840,
  "num_train_epochs": 40,
  "total_flos": 1.84657635477504e+19,
  "trial_name": null,
  "trial_params": null
}