trainer_state.json

{
  "best_metric": 0.916,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0443/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": Infinity,
      "learning_rate": 0.0004287859287859288,
      "loss": 0.7658,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7933333333333333,
      "eval_loss": 0.7393925786018372,
      "eval_runtime": 8.1221,
      "eval_samples_per_second": 461.706,
      "eval_steps_per_second": 7.264,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.8474286794662476,
      "learning_rate": 0.00035735735735735734,
      "loss": 0.3929,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8386666666666667,
      "eval_loss": 0.5318993926048279,
      "eval_runtime": 7.5637,
      "eval_samples_per_second": 495.789,
      "eval_steps_per_second": 7.8,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.4844480752944946,
      "learning_rate": 0.00028592878592878594,
      "loss": 0.2291,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8528,
      "eval_loss": 0.5035538673400879,
      "eval_runtime": 8.2616,
      "eval_samples_per_second": 453.907,
      "eval_steps_per_second": 7.141,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.008696038275957108,
      "learning_rate": 0.0002145002145002145,
      "loss": 0.1252,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8869333333333334,
      "eval_loss": 0.4347723424434662,
      "eval_runtime": 7.1315,
      "eval_samples_per_second": 525.833,
      "eval_steps_per_second": 8.273,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.014719804748892784,
      "learning_rate": 0.00014307164307164307,
      "loss": 0.0719,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8944,
      "eval_loss": 0.3983060419559479,
      "eval_runtime": 7.1205,
      "eval_samples_per_second": 526.651,
      "eval_steps_per_second": 8.286,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.01293383352458477,
      "learning_rate": 7.164307164307165e-05,
      "loss": 0.0208,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9112,
      "eval_loss": 0.3664200007915497,
      "eval_runtime": 7.0718,
      "eval_samples_per_second": 530.272,
      "eval_steps_per_second": 8.343,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.013020656071603298,
      "learning_rate": 2.1450021450021452e-07,
      "loss": 0.0057,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.916,
      "eval_loss": 0.3490810692310333,
      "eval_runtime": 6.9961,
      "eval_samples_per_second": 536.011,
      "eval_steps_per_second": 8.433,
      "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
}