trainer_state.json

{
  "best_metric": 0.8226666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0737/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": 0.0005,
      "loss": 0.8639,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8154666666666667,
      "eval_loss": 0.6495873332023621,
      "eval_runtime": 11.2121,
      "eval_samples_per_second": 334.46,
      "eval_steps_per_second": 5.262,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 11.718376159667969,
      "learning_rate": 0.0005,
      "loss": 0.5306,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8112,
      "eval_loss": 0.6732746958732605,
      "eval_runtime": 10.7531,
      "eval_samples_per_second": 348.737,
      "eval_steps_per_second": 5.487,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.493921279907227,
      "learning_rate": 0.0005,
      "loss": 0.4044,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7834666666666666,
      "eval_loss": 0.7679876685142517,
      "eval_runtime": 10.9151,
      "eval_samples_per_second": 343.56,
      "eval_steps_per_second": 5.405,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 17.04576873779297,
      "learning_rate": 0.0005,
      "loss": 0.3494,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7890666666666667,
      "eval_loss": 0.8043601512908936,
      "eval_runtime": 10.8528,
      "eval_samples_per_second": 345.532,
      "eval_steps_per_second": 5.436,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 31.484331130981445,
      "learning_rate": 0.0005,
      "loss": 0.2873,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8026666666666666,
      "eval_loss": 0.7484498620033264,
      "eval_runtime": 10.9869,
      "eval_samples_per_second": 341.315,
      "eval_steps_per_second": 5.37,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.6981940269470215,
      "learning_rate": 0.0005,
      "loss": 0.2624,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8061333333333334,
      "eval_loss": 0.7534906268119812,
      "eval_runtime": 10.8701,
      "eval_samples_per_second": 344.983,
      "eval_steps_per_second": 5.428,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.0058095999993383884,
      "learning_rate": 0.0005,
      "loss": 0.2336,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8226666666666667,
      "eval_loss": 0.7334408760070801,
      "eval_runtime": 10.858,
      "eval_samples_per_second": 345.366,
      "eval_steps_per_second": 5.434,
      "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
}