trainer_state.json

{
  "best_metric": 0.8714666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0208/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 481.89093017578125,
      "learning_rate": 8.571428571428571e-05,
      "loss": 3.0034,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6834666666666667,
      "eval_loss": 1.2740432024002075,
      "eval_runtime": 7.4412,
      "eval_samples_per_second": 503.948,
      "eval_steps_per_second": 7.929,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 46.825584411621094,
      "learning_rate": 7.142857142857143e-05,
      "loss": 0.8607,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7904,
      "eval_loss": 0.6848674416542053,
      "eval_runtime": 7.587,
      "eval_samples_per_second": 494.265,
      "eval_steps_per_second": 7.776,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 95.84591674804688,
      "learning_rate": 5.714285714285714e-05,
      "loss": 0.5388,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8373333333333334,
      "eval_loss": 0.5212387442588806,
      "eval_runtime": 7.4367,
      "eval_samples_per_second": 504.256,
      "eval_steps_per_second": 7.934,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 50.84409713745117,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.4069,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8610666666666666,
      "eval_loss": 0.46471261978149414,
      "eval_runtime": 7.5806,
      "eval_samples_per_second": 494.686,
      "eval_steps_per_second": 7.783,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 130.2327117919922,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.3413,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8565333333333334,
      "eval_loss": 0.4766438901424408,
      "eval_runtime": 7.4415,
      "eval_samples_per_second": 503.931,
      "eval_steps_per_second": 7.929,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 47.79440689086914,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.2873,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8714666666666666,
      "eval_loss": 0.42603567242622375,
      "eval_runtime": 7.4248,
      "eval_samples_per_second": 505.065,
      "eval_steps_per_second": 7.946,
      "step": 1998
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.9058575437312e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}