trainer_state.json

{
  "best_metric": 0.8757333333333334,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0774/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 84.29115295410156,
      "learning_rate": 8.728616793536588e-05,
      "loss": 2.8639,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7216,
      "eval_loss": 1.081047773361206,
      "eval_runtime": 13.275,
      "eval_samples_per_second": 282.485,
      "eval_steps_per_second": 4.444,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 73.08370208740234,
      "learning_rate": 7.947199994035401e-05,
      "loss": 0.7708,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8234666666666667,
      "eval_loss": 0.582588255405426,
      "eval_runtime": 13.6166,
      "eval_samples_per_second": 275.399,
      "eval_steps_per_second": 4.333,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 65.30943298339844,
      "learning_rate": 6.75e-05,
      "loss": 0.4751,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8552,
      "eval_loss": 0.47787266969680786,
      "eval_runtime": 13.208,
      "eval_samples_per_second": 283.918,
      "eval_steps_per_second": 4.467,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 71.91437530517578,
      "learning_rate": 5.2814167995011876e-05,
      "loss": 0.3353,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8656,
      "eval_loss": 0.4331841766834259,
      "eval_runtime": 13.2828,
      "eval_samples_per_second": 282.321,
      "eval_steps_per_second": 4.442,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 67.32086944580078,
      "learning_rate": 3.7185832004988136e-05,
      "loss": 0.246,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8656,
      "eval_loss": 0.44026950001716614,
      "eval_runtime": 13.1614,
      "eval_samples_per_second": 284.923,
      "eval_steps_per_second": 4.483,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 107.78864288330078,
      "learning_rate": 2.250000000000001e-05,
      "loss": 0.1888,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8749333333333333,
      "eval_loss": 0.4184397757053375,
      "eval_runtime": 13.5493,
      "eval_samples_per_second": 276.766,
      "eval_steps_per_second": 4.354,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 101.49785614013672,
      "learning_rate": 1.0528000059645995e-05,
      "loss": 0.1586,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8728,
      "eval_loss": 0.42355450987815857,
      "eval_runtime": 13.0709,
      "eval_samples_per_second": 286.898,
      "eval_steps_per_second": 4.514,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 154.8398895263672,
      "learning_rate": 2.7138320646341257e-06,
      "loss": 0.1391,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8757333333333334,
      "eval_loss": 0.42445671558380127,
      "eval_runtime": 13.1729,
      "eval_samples_per_second": 284.675,
      "eval_steps_per_second": 4.479,
      "step": 2664
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6.5411433916416e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}