trainer_state.json

{
  "best_metric": 0.932,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0344/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": 1.2888984680175781,
      "learning_rate": 0.0002851453301853628,
      "loss": 0.5965,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8690666666666667,
      "eval_loss": 0.465873658657074,
      "eval_runtime": 7.0703,
      "eval_samples_per_second": 530.387,
      "eval_steps_per_second": 8.345,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.08037152141332626,
      "learning_rate": 0.00024352347027881003,
      "loss": 0.2635,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8797333333333334,
      "eval_loss": 0.41160324215888977,
      "eval_runtime": 6.9287,
      "eval_samples_per_second": 541.226,
      "eval_steps_per_second": 8.515,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.835471153259277,
      "learning_rate": 0.00018337814009344714,
      "loss": 0.1401,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.896,
      "eval_loss": 0.385769784450531,
      "eval_runtime": 6.4106,
      "eval_samples_per_second": 584.971,
      "eval_steps_per_second": 9.204,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.7861754298210144,
      "learning_rate": 0.00011662185990655284,
      "loss": 0.0625,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9170666666666667,
      "eval_loss": 0.33513620495796204,
      "eval_runtime": 6.2066,
      "eval_samples_per_second": 604.197,
      "eval_steps_per_second": 9.506,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.014386146329343319,
      "learning_rate": 5.6476529721189974e-05,
      "loss": 0.016,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9232,
      "eval_loss": 0.31964394450187683,
      "eval_runtime": 6.213,
      "eval_samples_per_second": 603.573,
      "eval_steps_per_second": 9.496,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0076508307829499245,
      "learning_rate": 1.4854669814637143e-05,
      "loss": 0.004,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9306666666666666,
      "eval_loss": 0.2881596088409424,
      "eval_runtime": 7.3941,
      "eval_samples_per_second": 507.162,
      "eval_steps_per_second": 7.979,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.3623114161753767e-10,
      "loss": 0.0017,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.932,
      "eval_loss": 0.2890087366104126,
      "eval_runtime": 6.3182,
      "eval_samples_per_second": 593.521,
      "eval_steps_per_second": 9.338,
      "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
}