trainer_state.json

{
  "best_metric": 0.9024,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0982/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": 50.51131820678711,
      "learning_rate": 0.00047524221697560476,
      "loss": 1.1982,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8408,
      "eval_loss": 0.5318765044212341,
      "eval_runtime": 7.1367,
      "eval_samples_per_second": 525.455,
      "eval_steps_per_second": 8.267,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 40.297908782958984,
      "learning_rate": 0.0004058724504646834,
      "loss": 0.3747,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8712,
      "eval_loss": 0.42327776551246643,
      "eval_runtime": 7.11,
      "eval_samples_per_second": 527.429,
      "eval_steps_per_second": 8.298,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 20.956676483154297,
      "learning_rate": 0.0003056302334890786,
      "loss": 0.1881,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8864,
      "eval_loss": 0.3870830833911896,
      "eval_runtime": 7.1183,
      "eval_samples_per_second": 526.808,
      "eval_steps_per_second": 8.288,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 23.286636352539062,
      "learning_rate": 0.00019436976651092142,
      "loss": 0.0924,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8930666666666667,
      "eval_loss": 0.3907255232334137,
      "eval_runtime": 7.0715,
      "eval_samples_per_second": 530.297,
      "eval_steps_per_second": 8.343,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 27.71915626525879,
      "learning_rate": 9.412754953531663e-05,
      "loss": 0.04,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.896,
      "eval_loss": 0.407990962266922,
      "eval_runtime": 7.0773,
      "eval_samples_per_second": 529.864,
      "eval_steps_per_second": 8.337,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 43.776275634765625,
      "learning_rate": 2.4757783024395242e-05,
      "loss": 0.0179,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9005333333333333,
      "eval_loss": 0.41833579540252686,
      "eval_runtime": 6.9855,
      "eval_samples_per_second": 536.827,
      "eval_steps_per_second": 8.446,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 52.20749282836914,
      "learning_rate": 0.0,
      "loss": 0.0156,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9024,
      "eval_loss": 0.4021703898906708,
      "eval_runtime": 7.4561,
      "eval_samples_per_second": 502.944,
      "eval_steps_per_second": 7.913,
      "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
}