trainer_state.json

{
  "best_metric": 0.9005333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0894/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": 27.92206382751465,
      "learning_rate": 0.00047524221697560476,
      "loss": 1.3904,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8138666666666666,
      "eval_loss": 0.615444540977478,
      "eval_runtime": 14.1657,
      "eval_samples_per_second": 264.723,
      "eval_steps_per_second": 4.165,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 22.145153045654297,
      "learning_rate": 0.0004058724504646834,
      "loss": 0.4625,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8621333333333333,
      "eval_loss": 0.46109023690223694,
      "eval_runtime": 14.1074,
      "eval_samples_per_second": 265.819,
      "eval_steps_per_second": 4.182,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 133.33372497558594,
      "learning_rate": 0.0003056302334890786,
      "loss": 0.2717,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8818666666666667,
      "eval_loss": 0.3966662883758545,
      "eval_runtime": 14.1611,
      "eval_samples_per_second": 264.81,
      "eval_steps_per_second": 4.166,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 25.279130935668945,
      "learning_rate": 0.00019436976651092142,
      "loss": 0.1491,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8858666666666667,
      "eval_loss": 0.4387665390968323,
      "eval_runtime": 14.1117,
      "eval_samples_per_second": 265.737,
      "eval_steps_per_second": 4.181,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 44.18183898925781,
      "learning_rate": 9.412754953531663e-05,
      "loss": 0.0918,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8968,
      "eval_loss": 0.3821743130683899,
      "eval_runtime": 14.1049,
      "eval_samples_per_second": 265.865,
      "eval_steps_per_second": 4.183,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 33.59433364868164,
      "learning_rate": 2.4757783024395242e-05,
      "loss": 0.0491,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.896,
      "eval_loss": 0.4001845419406891,
      "eval_runtime": 14.1364,
      "eval_samples_per_second": 265.273,
      "eval_steps_per_second": 4.174,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 26.620798110961914,
      "learning_rate": 0.0,
      "loss": 0.0372,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9005333333333333,
      "eval_loss": 0.3991623818874359,
      "eval_runtime": 15.9548,
      "eval_samples_per_second": 235.039,
      "eval_steps_per_second": 3.698,
      "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
}