trainer_state.json

{
  "best_metric": 0.9301333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0899/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": 0.4434361755847931,
      "learning_rate": 0.0002571428571428571,
      "loss": 0.6598,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.872,
      "eval_loss": 0.41814151406288147,
      "eval_runtime": 6.3264,
      "eval_samples_per_second": 592.756,
      "eval_steps_per_second": 9.326,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7524564266204834,
      "learning_rate": 0.00021428571428571427,
      "loss": 0.2831,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8824,
      "eval_loss": 0.4097920060157776,
      "eval_runtime": 6.4382,
      "eval_samples_per_second": 582.463,
      "eval_steps_per_second": 9.164,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.035201478749513626,
      "learning_rate": 0.0001714285714285714,
      "loss": 0.1566,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8944,
      "eval_loss": 0.3915570378303528,
      "eval_runtime": 6.3367,
      "eval_samples_per_second": 591.791,
      "eval_steps_per_second": 9.311,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.10698332637548447,
      "learning_rate": 0.00012857142857142855,
      "loss": 0.0805,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9048,
      "eval_loss": 0.37031450867652893,
      "eval_runtime": 6.042,
      "eval_samples_per_second": 620.658,
      "eval_steps_per_second": 9.765,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.6519839763641357,
      "learning_rate": 8.57142857142857e-05,
      "loss": 0.0366,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9165333333333333,
      "eval_loss": 0.3408195674419403,
      "eval_runtime": 6.3523,
      "eval_samples_per_second": 590.338,
      "eval_steps_per_second": 9.288,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.005406931508332491,
      "learning_rate": 4.285714285714285e-05,
      "loss": 0.014,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9272,
      "eval_loss": 0.31809237599372864,
      "eval_runtime": 6.0469,
      "eval_samples_per_second": 620.152,
      "eval_steps_per_second": 9.757,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.2870012870012868e-07,
      "loss": 0.005,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9301333333333334,
      "eval_loss": 0.30823466181755066,
      "eval_runtime": 6.9304,
      "eval_samples_per_second": 541.094,
      "eval_steps_per_second": 8.513,
      "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
}