trainer_state.json

{
  "best_metric": 0.9037333333333334,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0365/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": 52.876800537109375,
      "learning_rate": 0.00047524221697560476,
      "loss": 1.2342,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8386666666666667,
      "eval_loss": 0.5541044473648071,
      "eval_runtime": 7.7572,
      "eval_samples_per_second": 483.421,
      "eval_steps_per_second": 7.606,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 32.929443359375,
      "learning_rate": 0.0004058724504646834,
      "loss": 0.406,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8704,
      "eval_loss": 0.4278021454811096,
      "eval_runtime": 7.7302,
      "eval_samples_per_second": 485.107,
      "eval_steps_per_second": 7.632,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 21.35298728942871,
      "learning_rate": 0.0003056302334890786,
      "loss": 0.2103,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8872,
      "eval_loss": 0.3988263010978699,
      "eval_runtime": 7.7185,
      "eval_samples_per_second": 485.845,
      "eval_steps_per_second": 7.644,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 27.241439819335938,
      "learning_rate": 0.00019436976651092142,
      "loss": 0.0997,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8954666666666666,
      "eval_loss": 0.4089593291282654,
      "eval_runtime": 7.7412,
      "eval_samples_per_second": 484.421,
      "eval_steps_per_second": 7.622,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 74.49508666992188,
      "learning_rate": 9.412754953531663e-05,
      "loss": 0.0509,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9005333333333333,
      "eval_loss": 0.38281920552253723,
      "eval_runtime": 8.1072,
      "eval_samples_per_second": 462.553,
      "eval_steps_per_second": 7.277,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 21.206335067749023,
      "learning_rate": 2.4757783024395242e-05,
      "loss": 0.0222,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9034666666666666,
      "eval_loss": 0.3864777386188507,
      "eval_runtime": 7.6781,
      "eval_samples_per_second": 488.402,
      "eval_steps_per_second": 7.684,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 29.021636962890625,
      "learning_rate": 0.0,
      "loss": 0.0147,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9037333333333334,
      "eval_loss": 0.38812726736068726,
      "eval_runtime": 8.4339,
      "eval_samples_per_second": 444.633,
      "eval_steps_per_second": 6.996,
      "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
}