trainer_state.json

{
  "best_metric": 0.8792,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0390/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": 59.93433380126953,
      "learning_rate": 0.0003,
      "loss": 1.477,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8184,
      "eval_loss": 0.57869952917099,
      "eval_runtime": 14.5623,
      "eval_samples_per_second": 257.514,
      "eval_steps_per_second": 4.052,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 27.24513053894043,
      "learning_rate": 0.0003,
      "loss": 0.4272,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8589333333333333,
      "eval_loss": 0.446659654378891,
      "eval_runtime": 14.4286,
      "eval_samples_per_second": 259.901,
      "eval_steps_per_second": 4.089,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 18.975833892822266,
      "learning_rate": 0.0003,
      "loss": 0.285,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8674666666666667,
      "eval_loss": 0.45267823338508606,
      "eval_runtime": 14.4237,
      "eval_samples_per_second": 259.989,
      "eval_steps_per_second": 4.09,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 29.45461654663086,
      "learning_rate": 0.0003,
      "loss": 0.1932,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8712,
      "eval_loss": 0.4472934901714325,
      "eval_runtime": 14.5608,
      "eval_samples_per_second": 257.541,
      "eval_steps_per_second": 4.052,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 38.71519470214844,
      "learning_rate": 0.0003,
      "loss": 0.1458,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8669333333333333,
      "eval_loss": 0.5022250413894653,
      "eval_runtime": 14.4348,
      "eval_samples_per_second": 259.789,
      "eval_steps_per_second": 4.087,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 32.20060348510742,
      "learning_rate": 0.0003,
      "loss": 0.1221,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8709333333333333,
      "eval_loss": 0.4980267584323883,
      "eval_runtime": 14.4064,
      "eval_samples_per_second": 260.302,
      "eval_steps_per_second": 4.095,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 31.531612396240234,
      "learning_rate": 0.0003,
      "loss": 0.0894,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8792,
      "eval_loss": 0.5044906139373779,
      "eval_runtime": 14.4267,
      "eval_samples_per_second": 259.934,
      "eval_steps_per_second": 4.09,
      "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
}