trainer_state.json

{
  "best_metric": 0.9370666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0691/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": 0.16396696865558624,
      "learning_rate": 8.554359905560886e-05,
      "loss": 0.6399,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9154666666666667,
      "eval_loss": 0.2827706038951874,
      "eval_runtime": 10.9906,
      "eval_samples_per_second": 341.201,
      "eval_steps_per_second": 5.368,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.07914759963750839,
      "learning_rate": 7.305704108364301e-05,
      "loss": 0.1638,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9224,
      "eval_loss": 0.2551507353782654,
      "eval_runtime": 11.507,
      "eval_samples_per_second": 325.888,
      "eval_steps_per_second": 5.127,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.027219807729125023,
      "learning_rate": 5.501344202803415e-05,
      "loss": 0.0828,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9309333333333333,
      "eval_loss": 0.2469257265329361,
      "eval_runtime": 10.9884,
      "eval_samples_per_second": 341.269,
      "eval_steps_per_second": 5.369,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.045479245483875275,
      "learning_rate": 3.498655797196586e-05,
      "loss": 0.037,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9306666666666666,
      "eval_loss": 0.2559111714363098,
      "eval_runtime": 11.0342,
      "eval_samples_per_second": 339.851,
      "eval_steps_per_second": 5.347,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.00950576364994049,
      "learning_rate": 1.6942958916356995e-05,
      "loss": 0.0126,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.932,
      "eval_loss": 0.2517668604850769,
      "eval_runtime": 11.3758,
      "eval_samples_per_second": 329.647,
      "eval_steps_per_second": 5.186,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.04000629857182503,
      "learning_rate": 4.456400944391144e-06,
      "loss": 0.0055,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9362666666666667,
      "eval_loss": 0.2548176348209381,
      "eval_runtime": 11.0725,
      "eval_samples_per_second": 338.678,
      "eval_steps_per_second": 5.329,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 4.086934248526131e-11,
      "loss": 0.0039,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9370666666666667,
      "eval_loss": 0.24709057807922363,
      "eval_runtime": 10.9983,
      "eval_samples_per_second": 340.961,
      "eval_steps_per_second": 5.364,
      "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
}