trainer_state.json

{
  "best_metric": 0.8805333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0999/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": 172.3248748779297,
      "learning_rate": 8.728616793536588e-05,
      "loss": 3.029,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6704,
      "eval_loss": 1.2995142936706543,
      "eval_runtime": 8.2122,
      "eval_samples_per_second": 456.636,
      "eval_steps_per_second": 7.184,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 105.65258026123047,
      "learning_rate": 7.947199994035401e-05,
      "loss": 0.8883,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8096,
      "eval_loss": 0.6433312892913818,
      "eval_runtime": 7.7999,
      "eval_samples_per_second": 480.778,
      "eval_steps_per_second": 7.564,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 77.92212677001953,
      "learning_rate": 6.75e-05,
      "loss": 0.5667,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8448,
      "eval_loss": 0.5035193562507629,
      "eval_runtime": 7.6185,
      "eval_samples_per_second": 492.222,
      "eval_steps_per_second": 7.744,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 56.192420959472656,
      "learning_rate": 5.2814167995011876e-05,
      "loss": 0.4345,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8522666666666666,
      "eval_loss": 0.4557854235172272,
      "eval_runtime": 7.6535,
      "eval_samples_per_second": 489.97,
      "eval_steps_per_second": 7.709,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 47.805389404296875,
      "learning_rate": 3.7185832004988136e-05,
      "loss": 0.3575,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8701333333333333,
      "eval_loss": 0.4113144278526306,
      "eval_runtime": 7.7203,
      "eval_samples_per_second": 485.733,
      "eval_steps_per_second": 7.642,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 38.57602310180664,
      "learning_rate": 2.250000000000001e-05,
      "loss": 0.3062,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8768,
      "eval_loss": 0.39695754647254944,
      "eval_runtime": 7.6307,
      "eval_samples_per_second": 491.438,
      "eval_steps_per_second": 7.732,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 63.736572265625,
      "learning_rate": 1.0528000059645995e-05,
      "loss": 0.2758,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.872,
      "eval_loss": 0.40313589572906494,
      "eval_runtime": 12.7797,
      "eval_samples_per_second": 293.434,
      "eval_steps_per_second": 4.617,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 38.75566101074219,
      "learning_rate": 2.7138320646341257e-06,
      "loss": 0.2546,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8805333333333333,
      "eval_loss": 0.39343005418777466,
      "eval_runtime": 7.6354,
      "eval_samples_per_second": 491.131,
      "eval_steps_per_second": 7.727,
      "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
}