trainer_state.json

{
  "best_metric": 0.856,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0200/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": 5e-05,
      "loss": 2.0622,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.748,
      "eval_loss": 0.9497717618942261,
      "eval_runtime": 14.4691,
      "eval_samples_per_second": 259.173,
      "eval_steps_per_second": 4.078,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 5e-05,
      "loss": 0.716,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8008,
      "eval_loss": 0.6426911950111389,
      "eval_runtime": 13.1437,
      "eval_samples_per_second": 285.309,
      "eval_steps_per_second": 4.489,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.2363955974578857,
      "learning_rate": 5e-05,
      "loss": 0.461,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.828,
      "eval_loss": 0.5611403584480286,
      "eval_runtime": 13.5167,
      "eval_samples_per_second": 277.435,
      "eval_steps_per_second": 4.365,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 15.931729316711426,
      "learning_rate": 5e-05,
      "loss": 0.3245,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8368,
      "eval_loss": 0.5382492542266846,
      "eval_runtime": 12.6294,
      "eval_samples_per_second": 296.927,
      "eval_steps_per_second": 4.672,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 42.53856658935547,
      "learning_rate": 5e-05,
      "loss": 0.2379,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8477333333333333,
      "eval_loss": 0.502540647983551,
      "eval_runtime": 13.5925,
      "eval_samples_per_second": 275.888,
      "eval_steps_per_second": 4.341,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": Infinity,
      "learning_rate": 5e-05,
      "loss": 0.1813,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8498666666666667,
      "eval_loss": 0.48619166016578674,
      "eval_runtime": 13.3205,
      "eval_samples_per_second": 281.521,
      "eval_steps_per_second": 4.429,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.3675918579101562,
      "learning_rate": 5e-05,
      "loss": 0.1314,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.856,
      "eval_loss": 0.48488086462020874,
      "eval_runtime": 13.4544,
      "eval_samples_per_second": 278.718,
      "eval_steps_per_second": 4.385,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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
}