trainer_state.json

{
  "best_metric": 0.8125333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0550/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": Infinity,
      "learning_rate": 0.0003,
      "loss": 1.7328,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7488,
      "eval_loss": 0.85762619972229,
      "eval_runtime": 5.8935,
      "eval_samples_per_second": 636.29,
      "eval_steps_per_second": 10.011,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.571508884429932,
      "learning_rate": 0.0003,
      "loss": 0.6873,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7829333333333334,
      "eval_loss": 0.7133203148841858,
      "eval_runtime": 5.8891,
      "eval_samples_per_second": 636.774,
      "eval_steps_per_second": 10.019,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.6891889572143555,
      "learning_rate": 0.0003,
      "loss": 0.4592,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8024,
      "eval_loss": 0.6652063727378845,
      "eval_runtime": 5.9098,
      "eval_samples_per_second": 634.537,
      "eval_steps_per_second": 9.983,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.861435413360596,
      "learning_rate": 0.0003,
      "loss": 0.3305,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8072,
      "eval_loss": 0.6639143228530884,
      "eval_runtime": 6.061,
      "eval_samples_per_second": 618.714,
      "eval_steps_per_second": 9.734,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.08786794543266296,
      "learning_rate": 0.0003,
      "loss": 0.2571,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8125333333333333,
      "eval_loss": 0.678558349609375,
      "eval_runtime": 6.0312,
      "eval_samples_per_second": 621.763,
      "eval_steps_per_second": 9.782,
      "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": 8.2370660311296e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}