trainer_state.json

{
  "best_metric": 0.884,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0295/checkpoints/checkpoint-1665",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1665,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 49.29814910888672,
      "learning_rate": 0.0001,
      "loss": 2.7267,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7568,
      "eval_loss": 0.9301578998565674,
      "eval_runtime": 16.7824,
      "eval_samples_per_second": 223.448,
      "eval_steps_per_second": 3.516,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 45.87650680541992,
      "learning_rate": 0.0001,
      "loss": 0.6739,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8408,
      "eval_loss": 0.5322536826133728,
      "eval_runtime": 17.3107,
      "eval_samples_per_second": 216.629,
      "eval_steps_per_second": 3.408,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 60.56796646118164,
      "learning_rate": 0.0001,
      "loss": 0.4239,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.86,
      "eval_loss": 0.4679814875125885,
      "eval_runtime": 17.2191,
      "eval_samples_per_second": 217.781,
      "eval_steps_per_second": 3.426,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 71.6979751586914,
      "learning_rate": 0.0001,
      "loss": 0.2919,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8816,
      "eval_loss": 0.42274728417396545,
      "eval_runtime": 17.4091,
      "eval_samples_per_second": 215.405,
      "eval_steps_per_second": 3.389,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 41.79677200317383,
      "learning_rate": 0.0001,
      "loss": 0.1973,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.884,
      "eval_loss": 0.4206501245498657,
      "eval_runtime": 16.1467,
      "eval_samples_per_second": 232.246,
      "eval_steps_per_second": 3.654,
      "step": 1665
    }
  ],
  "logging_steps": 500,
  "max_steps": 1665,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.088214619776e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}