trainer_state.json

{
  "best_metric": 0.8434666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0043/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "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": 9e-05,
      "loss": 1.5022,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7904,
      "eval_loss": 0.7341175675392151,
      "eval_runtime": 11.6022,
      "eval_samples_per_second": 323.214,
      "eval_steps_per_second": 5.085,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.42823654413223267,
      "learning_rate": 9e-05,
      "loss": 0.4888,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8256,
      "eval_loss": 0.588905394077301,
      "eval_runtime": 12.3698,
      "eval_samples_per_second": 303.158,
      "eval_steps_per_second": 4.77,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 7.229664325714111,
      "learning_rate": 9e-05,
      "loss": 0.2769,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8354666666666667,
      "eval_loss": 0.5504579544067383,
      "eval_runtime": 12.2867,
      "eval_samples_per_second": 305.209,
      "eval_steps_per_second": 4.802,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.08456508815288544,
      "learning_rate": 9e-05,
      "loss": 0.1766,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.84,
      "eval_loss": 0.5504480004310608,
      "eval_runtime": 11.6867,
      "eval_samples_per_second": 320.878,
      "eval_steps_per_second": 5.048,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.2784363925457001,
      "learning_rate": 9e-05,
      "loss": 0.1205,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8378666666666666,
      "eval_loss": 0.577860414981842,
      "eval_runtime": 11.7772,
      "eval_samples_per_second": 318.411,
      "eval_steps_per_second": 5.01,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.201807975769043,
      "learning_rate": 9e-05,
      "loss": 0.0857,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8434666666666667,
      "eval_loss": 0.6023405194282532,
      "eval_runtime": 11.6552,
      "eval_samples_per_second": 321.745,
      "eval_steps_per_second": 5.062,
      "step": 1998
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}