trainer_state.json

{
  "best_metric": 0.9645333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0235/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": 0.07281150668859482,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.5624,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9442666666666667,
      "eval_loss": 0.20104873180389404,
      "eval_runtime": 7.0508,
      "eval_samples_per_second": 531.853,
      "eval_steps_per_second": 8.368,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.1520155817270279,
      "learning_rate": 8.535533905932738e-05,
      "loss": 0.101,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9397333333333333,
      "eval_loss": 0.21073482930660248,
      "eval_runtime": 5.7529,
      "eval_samples_per_second": 651.85,
      "eval_steps_per_second": 10.256,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.03500928729772568,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0438,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.948,
      "eval_loss": 0.19289866089820862,
      "eval_runtime": 7.3186,
      "eval_samples_per_second": 512.395,
      "eval_steps_per_second": 8.062,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.027462273836135864,
      "learning_rate": 5e-05,
      "loss": 0.0141,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9536,
      "eval_loss": 0.19727133214473724,
      "eval_runtime": 5.6175,
      "eval_samples_per_second": 667.554,
      "eval_steps_per_second": 10.503,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.009924763813614845,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0046,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9605333333333334,
      "eval_loss": 0.172583669424057,
      "eval_runtime": 5.8095,
      "eval_samples_per_second": 645.495,
      "eval_steps_per_second": 10.156,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.006727916654199362,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.0014,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9634666666666667,
      "eval_loss": 0.16203773021697998,
      "eval_runtime": 5.7633,
      "eval_samples_per_second": 650.665,
      "eval_steps_per_second": 10.237,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 3.82861996937452e-06,
      "loss": 0.0009,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9645333333333334,
      "eval_loss": 0.15968413650989532,
      "eval_runtime": 5.6872,
      "eval_samples_per_second": 659.38,
      "eval_steps_per_second": 10.374,
      "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
}