trainer_state.json

{
  "best_metric": 0.7642666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0533/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": 158.14614868164062,
      "learning_rate": 2.5714285714285714e-05,
      "loss": 3.8632,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.26986666666666664,
      "eval_loss": 3.7640650272369385,
      "eval_runtime": 13.2541,
      "eval_samples_per_second": 282.93,
      "eval_steps_per_second": 4.451,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 86.85962677001953,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 3.4428,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5024,
      "eval_loss": 3.081892728805542,
      "eval_runtime": 13.1111,
      "eval_samples_per_second": 286.018,
      "eval_steps_per_second": 4.5,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 159.7128448486328,
      "learning_rate": 1.7142857142857142e-05,
      "loss": 2.435,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6402666666666667,
      "eval_loss": 1.9003947973251343,
      "eval_runtime": 13.2336,
      "eval_samples_per_second": 283.37,
      "eval_steps_per_second": 4.458,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 412.1076354980469,
      "learning_rate": 1.2857142857142857e-05,
      "loss": 1.5582,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7042666666666667,
      "eval_loss": 1.3699098825454712,
      "eval_runtime": 13.3936,
      "eval_samples_per_second": 279.984,
      "eval_steps_per_second": 4.405,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 279.7771301269531,
      "learning_rate": 8.571428571428571e-06,
      "loss": 1.1942,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7352,
      "eval_loss": 1.1556174755096436,
      "eval_runtime": 13.2841,
      "eval_samples_per_second": 282.292,
      "eval_steps_per_second": 4.441,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 114.0682373046875,
      "learning_rate": 4.2857142857142855e-06,
      "loss": 1.021,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7514666666666666,
      "eval_loss": 0.9642806649208069,
      "eval_runtime": 13.0991,
      "eval_samples_per_second": 286.279,
      "eval_steps_per_second": 4.504,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 281.4178771972656,
      "learning_rate": 0.0,
      "loss": 0.9626,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7642666666666666,
      "eval_loss": 1.0092110633850098,
      "eval_runtime": 13.567,
      "eval_samples_per_second": 276.405,
      "eval_steps_per_second": 4.349,
      "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": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}