trainer_state.json

{
  "best_metric": 0.8797333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0715/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": 6.003003003003003e-05,
      "loss": 1.6458,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7882666666666667,
      "eval_loss": 0.7441364526748657,
      "eval_runtime": 10.7733,
      "eval_samples_per_second": 348.084,
      "eval_steps_per_second": 5.477,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.7071380615234375,
      "learning_rate": 5.0030030030030024e-05,
      "loss": 0.5435,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8269333333333333,
      "eval_loss": 0.5719521045684814,
      "eval_runtime": 10.7975,
      "eval_samples_per_second": 347.303,
      "eval_steps_per_second": 5.464,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.2586959600448608,
      "learning_rate": 4.003003003003003e-05,
      "loss": 0.3106,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8546666666666667,
      "eval_loss": 0.48576465249061584,
      "eval_runtime": 10.7624,
      "eval_samples_per_second": 348.435,
      "eval_steps_per_second": 5.482,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.25216817855835,
      "learning_rate": 3.003003003003003e-05,
      "loss": 0.1806,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8632,
      "eval_loss": 0.45279714465141296,
      "eval_runtime": 10.8745,
      "eval_samples_per_second": 344.844,
      "eval_steps_per_second": 5.426,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.9892401099205017,
      "learning_rate": 2.003003003003003e-05,
      "loss": 0.1054,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.872,
      "eval_loss": 0.4356541633605957,
      "eval_runtime": 11.247,
      "eval_samples_per_second": 333.421,
      "eval_steps_per_second": 5.246,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.346026748418808,
      "learning_rate": 1.003003003003003e-05,
      "loss": 0.0567,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8781333333333333,
      "eval_loss": 0.43072709441185,
      "eval_runtime": 10.5046,
      "eval_samples_per_second": 356.985,
      "eval_steps_per_second": 5.617,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.2364611625671387,
      "learning_rate": 3.0030030030030026e-08,
      "loss": 0.0285,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8797333333333334,
      "eval_loss": 0.42263805866241455,
      "eval_runtime": 10.5985,
      "eval_samples_per_second": 353.824,
      "eval_steps_per_second": 5.567,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}