trainer_state.json

{
  "best_metric": 0.9464,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0127/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": 2.9100759384327822e-05,
      "loss": 1.1553,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9309333333333333,
      "eval_loss": 0.2940164804458618,
      "eval_runtime": 7.1293,
      "eval_samples_per_second": 525.997,
      "eval_steps_per_second": 8.276,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6072393655776978,
      "learning_rate": 2.6500767323004636e-05,
      "loss": 0.1665,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9293333333333333,
      "eval_loss": 0.23465624451637268,
      "eval_runtime": 6.61,
      "eval_samples_per_second": 567.32,
      "eval_steps_per_second": 8.926,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.16311264038085938,
      "learning_rate": 2.25136129892785e-05,
      "loss": 0.0672,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9373333333333334,
      "eval_loss": 0.21565885841846466,
      "eval_runtime": 6.5298,
      "eval_samples_per_second": 574.289,
      "eval_steps_per_second": 9.035,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.42402276396751404,
      "learning_rate": 1.7620206039927677e-05,
      "loss": 0.0266,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9426666666666667,
      "eval_loss": 0.21543291211128235,
      "eval_runtime": 6.2415,
      "eval_samples_per_second": 600.814,
      "eval_steps_per_second": 9.453,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.10640875995159149,
      "learning_rate": 1.241076357203891e-05,
      "loss": 0.0104,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9456,
      "eval_loss": 0.2163020521402359,
      "eval_runtime": 6.3291,
      "eval_samples_per_second": 592.499,
      "eval_steps_per_second": 9.322,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.029114890843629837,
      "learning_rate": 7.513621230422123e-06,
      "loss": 0.0055,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9461333333333334,
      "eval_loss": 0.21356888115406036,
      "eval_runtime": 6.5235,
      "eval_samples_per_second": 574.845,
      "eval_steps_per_second": 9.044,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.034476835280656815,
      "learning_rate": 3.519446655599853e-06,
      "loss": 0.0035,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9464,
      "eval_loss": 0.22093935310840607,
      "eval_runtime": 7.3336,
      "eval_samples_per_second": 511.347,
      "eval_steps_per_second": 8.045,
      "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
}