trainer_state.json

{
  "best_metric": 0.8752,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0970/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": 130.5298309326172,
      "learning_rate": 0.0001,
      "loss": 2.6282,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7621333333333333,
      "eval_loss": 0.845592200756073,
      "eval_runtime": 14.4425,
      "eval_samples_per_second": 259.65,
      "eval_steps_per_second": 4.085,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 154.12332153320312,
      "learning_rate": 0.0001,
      "loss": 0.6342,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8488,
      "eval_loss": 0.4909805357456207,
      "eval_runtime": 14.7049,
      "eval_samples_per_second": 255.017,
      "eval_steps_per_second": 4.012,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 57.6883430480957,
      "learning_rate": 0.0001,
      "loss": 0.3923,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8752,
      "eval_loss": 0.4054300785064697,
      "eval_runtime": 13.7866,
      "eval_samples_per_second": 272.002,
      "eval_steps_per_second": 4.28,
      "step": 999
    }
  ],
  "logging_steps": 500,
  "max_steps": 1332,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "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
}