trainer_state.json

{
  "best_metric": 0.9472,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0012/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": 0.4767933785915375,
      "learning_rate": 3e-05,
      "loss": 1.1904,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9354666666666667,
      "eval_loss": 0.29159611463546753,
      "eval_runtime": 13.2708,
      "eval_samples_per_second": 282.576,
      "eval_steps_per_second": 4.446,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.25969555974006653,
      "learning_rate": 3e-05,
      "loss": 0.156,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9376,
      "eval_loss": 0.21004101634025574,
      "eval_runtime": 12.5279,
      "eval_samples_per_second": 299.331,
      "eval_steps_per_second": 4.709,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.3778182864189148,
      "learning_rate": 3e-05,
      "loss": 0.0533,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9346666666666666,
      "eval_loss": 0.2261248677968979,
      "eval_runtime": 12.3941,
      "eval_samples_per_second": 302.563,
      "eval_steps_per_second": 4.76,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.08102396875619888,
      "learning_rate": 3e-05,
      "loss": 0.0193,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9416,
      "eval_loss": 0.22569406032562256,
      "eval_runtime": 13.8693,
      "eval_samples_per_second": 270.381,
      "eval_steps_per_second": 4.254,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.040736936032772064,
      "learning_rate": 3e-05,
      "loss": 0.0089,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9421333333333334,
      "eval_loss": 0.22167590260505676,
      "eval_runtime": 11.992,
      "eval_samples_per_second": 312.708,
      "eval_steps_per_second": 4.92,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.022913649678230286,
      "learning_rate": 3e-05,
      "loss": 0.0044,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9458666666666666,
      "eval_loss": 0.22502802312374115,
      "eval_runtime": 11.7701,
      "eval_samples_per_second": 318.605,
      "eval_steps_per_second": 5.013,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 3e-05,
      "loss": 0.0027,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9472,
      "eval_loss": 0.22873690724372864,
      "eval_runtime": 12.1994,
      "eval_samples_per_second": 307.393,
      "eval_steps_per_second": 4.836,
      "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
}