trainer_state.json

{
  "best_metric": 0.872,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0214/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "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": 2.886495269365963e-05,
      "loss": 2.4616,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7117333333333333,
      "eval_loss": 1.3081082105636597,
      "eval_runtime": 6.9809,
      "eval_samples_per_second": 537.177,
      "eval_steps_per_second": 8.452,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 2.5631588426322576e-05,
      "loss": 0.886,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8088,
      "eval_loss": 0.7198894619941711,
      "eval_runtime": 6.8155,
      "eval_samples_per_second": 550.214,
      "eval_steps_per_second": 8.657,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 65.33013153076172,
      "learning_rate": 2.0772920772568672e-05,
      "loss": 0.5043,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8397333333333333,
      "eval_loss": 0.5730645656585693,
      "eval_runtime": 6.6828,
      "eval_samples_per_second": 561.14,
      "eval_steps_per_second": 8.829,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.3731517791748047,
      "learning_rate": 1.5035378262848308e-05,
      "loss": 0.3216,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8541333333333333,
      "eval_loss": 0.5039607286453247,
      "eval_runtime": 7.1209,
      "eval_samples_per_second": 526.619,
      "eval_steps_per_second": 8.285,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.765988290309906,
      "learning_rate": 9.292449733314041e-06,
      "loss": 0.2131,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8632,
      "eval_loss": 0.4703887403011322,
      "eval_runtime": 6.7122,
      "eval_samples_per_second": 558.687,
      "eval_steps_per_second": 8.79,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.1489320993423462,
      "learning_rate": 4.418443992810701e-06,
      "loss": 0.1414,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8690666666666667,
      "eval_loss": 0.4573669731616974,
      "eval_runtime": 6.6707,
      "eval_samples_per_second": 562.157,
      "eval_steps_per_second": 8.845,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 5.046769142150879,
      "learning_rate": 1.1553842323578667e-06,
      "loss": 0.1021,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.872,
      "eval_loss": 0.44202059507369995,
      "eval_runtime": 7.3867,
      "eval_samples_per_second": 507.666,
      "eval_steps_per_second": 7.987,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}