trainer_state.json

{
  "best_metric": 0.8989333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0083/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": 7.101415634155273,
      "learning_rate": 0.00047524221697560476,
      "loss": 0.8242,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7754666666666666,
      "eval_loss": 0.7520184516906738,
      "eval_runtime": 14.6508,
      "eval_samples_per_second": 255.959,
      "eval_steps_per_second": 4.027,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.04289412871003151,
      "learning_rate": 0.0004058724504646834,
      "loss": 0.4463,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8290666666666666,
      "eval_loss": 0.5891080498695374,
      "eval_runtime": 11.3889,
      "eval_samples_per_second": 329.268,
      "eval_steps_per_second": 5.18,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.03138558194041252,
      "learning_rate": 0.0003056302334890786,
      "loss": 0.2568,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.852,
      "eval_loss": 0.4912415146827698,
      "eval_runtime": 11.7699,
      "eval_samples_per_second": 318.611,
      "eval_steps_per_second": 5.013,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.003731026779860258,
      "learning_rate": 0.00019436976651092142,
      "loss": 0.1193,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8658666666666667,
      "eval_loss": 0.4882695972919464,
      "eval_runtime": 11.5922,
      "eval_samples_per_second": 323.493,
      "eval_steps_per_second": 5.09,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.0440542697906494,
      "learning_rate": 9.412754953531663e-05,
      "loss": 0.0391,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8890666666666667,
      "eval_loss": 0.43462273478507996,
      "eval_runtime": 11.1036,
      "eval_samples_per_second": 337.728,
      "eval_steps_per_second": 5.314,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.39598193764686584,
      "learning_rate": 2.4757783024395242e-05,
      "loss": 0.0071,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8968,
      "eval_loss": 0.3908087909221649,
      "eval_runtime": 11.1134,
      "eval_samples_per_second": 337.432,
      "eval_steps_per_second": 5.309,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 2.2705190269589615e-10,
      "loss": 0.0023,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8989333333333334,
      "eval_loss": 0.3919280469417572,
      "eval_runtime": 10.8558,
      "eval_samples_per_second": 345.436,
      "eval_steps_per_second": 5.435,
      "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
}