trainer_state.json

{
  "best_metric": 0.9064,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0924/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": 0.33598026633262634,
      "learning_rate": 0.0004809698831278217,
      "loss": 0.8017,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8165333333333333,
      "eval_loss": 0.6317250728607178,
      "eval_runtime": 10.2258,
      "eval_samples_per_second": 366.719,
      "eval_steps_per_second": 5.77,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.06094444543123245,
      "learning_rate": 0.00042677669529663686,
      "loss": 0.4195,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8261333333333334,
      "eval_loss": 0.59303879737854,
      "eval_runtime": 10.2949,
      "eval_samples_per_second": 364.257,
      "eval_steps_per_second": 5.731,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.01121739111840725,
      "learning_rate": 0.0003456708580912725,
      "loss": 0.2571,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8426666666666667,
      "eval_loss": 0.5838651657104492,
      "eval_runtime": 10.8111,
      "eval_samples_per_second": 346.864,
      "eval_steps_per_second": 5.457,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.04552498459815979,
      "learning_rate": 0.00025,
      "loss": 0.1384,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8509333333333333,
      "eval_loss": 0.5807331800460815,
      "eval_runtime": 10.4208,
      "eval_samples_per_second": 359.857,
      "eval_steps_per_second": 5.662,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.010106077417731285,
      "learning_rate": 0.00015432914190872756,
      "loss": 0.0529,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8757333333333334,
      "eval_loss": 0.529794454574585,
      "eval_runtime": 10.2247,
      "eval_samples_per_second": 366.76,
      "eval_steps_per_second": 5.77,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0036816424690186977,
      "learning_rate": 7.322330470336314e-05,
      "loss": 0.0133,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9050666666666667,
      "eval_loss": 0.4096379578113556,
      "eval_runtime": 10.5064,
      "eval_samples_per_second": 356.925,
      "eval_steps_per_second": 5.616,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.9143099846872597e-05,
      "loss": 0.002,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9064,
      "eval_loss": 0.39579468965530396,
      "eval_runtime": 10.2361,
      "eval_samples_per_second": 366.352,
      "eval_steps_per_second": 5.764,
      "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
}