trainer_state.json

{
  "best_metric": 0.7794666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0196/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": 2.4192,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5829333333333333,
      "eval_loss": 1.489530086517334,
      "eval_runtime": 7.7118,
      "eval_samples_per_second": 486.267,
      "eval_steps_per_second": 7.651,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0003,
      "loss": 1.0183,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.696,
      "eval_loss": 1.014305591583252,
      "eval_runtime": 9.7517,
      "eval_samples_per_second": 384.547,
      "eval_steps_per_second": 6.05,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0003,
      "loss": 0.6702,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7488,
      "eval_loss": 0.8392720222473145,
      "eval_runtime": 8.5301,
      "eval_samples_per_second": 439.619,
      "eval_steps_per_second": 6.917,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 32.41567611694336,
      "learning_rate": 0.0003,
      "loss": 0.4855,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7384,
      "eval_loss": 0.8888546824455261,
      "eval_runtime": 8.0604,
      "eval_samples_per_second": 465.236,
      "eval_steps_per_second": 7.32,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 9.422261238098145,
      "learning_rate": 0.0003,
      "loss": 0.3738,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7794666666666666,
      "eval_loss": 0.7742811441421509,
      "eval_runtime": 8.8769,
      "eval_samples_per_second": 422.443,
      "eval_steps_per_second": 6.646,
      "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
}