trainer_state.json

{
  "best_metric": 0.8621333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0312/checkpoints/checkpoint-999",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 999,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 134.4685516357422,
      "learning_rate": 0.0003,
      "loss": 1.5342,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8133333333333334,
      "eval_loss": 0.5586355328559875,
      "eval_runtime": 12.7138,
      "eval_samples_per_second": 294.955,
      "eval_steps_per_second": 4.641,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 51.908111572265625,
      "learning_rate": 0.0003,
      "loss": 0.4571,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8464,
      "eval_loss": 0.4937789738178253,
      "eval_runtime": 12.7343,
      "eval_samples_per_second": 294.479,
      "eval_steps_per_second": 4.633,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 59.61941909790039,
      "learning_rate": 0.0003,
      "loss": 0.2996,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8621333333333333,
      "eval_loss": 0.4534153342247009,
      "eval_runtime": 12.5411,
      "eval_samples_per_second": 299.016,
      "eval_steps_per_second": 4.705,
      "step": 999
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.4529287718656e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}