trainer_state.json

{
  "best_metric": 0.9565333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0752/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": 0.2076968103647232,
      "learning_rate": 4.8096988312782174e-05,
      "loss": 0.7956,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9304,
      "eval_loss": 0.24332672357559204,
      "eval_runtime": 4.3729,
      "eval_samples_per_second": 857.555,
      "eval_steps_per_second": 13.492,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.39684343338012695,
      "learning_rate": 4.267766952966369e-05,
      "loss": 0.1105,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9437333333333333,
      "eval_loss": 0.19446074962615967,
      "eval_runtime": 3.9423,
      "eval_samples_per_second": 951.212,
      "eval_steps_per_second": 14.966,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.597532749176025,
      "learning_rate": 3.456708580912725e-05,
      "loss": 0.0359,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9506666666666667,
      "eval_loss": 0.18241624534130096,
      "eval_runtime": 3.9087,
      "eval_samples_per_second": 959.389,
      "eval_steps_per_second": 15.094,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.022740399464964867,
      "learning_rate": 2.5e-05,
      "loss": 0.0113,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9533333333333334,
      "eval_loss": 0.1838407814502716,
      "eval_runtime": 3.9488,
      "eval_samples_per_second": 949.647,
      "eval_steps_per_second": 14.941,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.031340811401605606,
      "learning_rate": 1.5432914190872757e-05,
      "loss": 0.0046,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9557333333333333,
      "eval_loss": 0.17339688539505005,
      "eval_runtime": 3.9447,
      "eval_samples_per_second": 950.654,
      "eval_steps_per_second": 14.957,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.05042155086994171,
      "learning_rate": 7.3223304703363135e-06,
      "loss": 0.0031,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9552,
      "eval_loss": 0.17329499125480652,
      "eval_runtime": 4.0489,
      "eval_samples_per_second": 926.168,
      "eval_steps_per_second": 14.572,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.91430998468726e-06,
      "loss": 0.002,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9565333333333333,
      "eval_loss": 0.17311204969882965,
      "eval_runtime": 5.8925,
      "eval_samples_per_second": 636.405,
      "eval_steps_per_second": 10.013,
      "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
}