trainer_state.json

{
  "best_metric": 0.8981333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0715/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": 19.785356521606445,
      "learning_rate": 0.0005,
      "loss": 1.2089,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8581333333333333,
      "eval_loss": 0.48082980513572693,
      "eval_runtime": 12.7874,
      "eval_samples_per_second": 293.258,
      "eval_steps_per_second": 4.614,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 35.12892150878906,
      "learning_rate": 0.0005,
      "loss": 0.4075,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8717333333333334,
      "eval_loss": 0.45028582215309143,
      "eval_runtime": 12.4155,
      "eval_samples_per_second": 302.041,
      "eval_steps_per_second": 4.752,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 24.936857223510742,
      "learning_rate": 0.0005,
      "loss": 0.2647,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.88,
      "eval_loss": 0.4236472547054291,
      "eval_runtime": 12.3929,
      "eval_samples_per_second": 302.593,
      "eval_steps_per_second": 4.761,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 13.124445915222168,
      "learning_rate": 0.0005,
      "loss": 0.1719,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8949333333333334,
      "eval_loss": 0.3917216658592224,
      "eval_runtime": 12.403,
      "eval_samples_per_second": 302.346,
      "eval_steps_per_second": 4.757,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 24.710926055908203,
      "learning_rate": 0.0005,
      "loss": 0.1411,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8925333333333333,
      "eval_loss": 0.39665472507476807,
      "eval_runtime": 13.476,
      "eval_samples_per_second": 278.272,
      "eval_steps_per_second": 4.378,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 17.258737564086914,
      "learning_rate": 0.0005,
      "loss": 0.1115,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8946666666666667,
      "eval_loss": 0.4111942648887634,
      "eval_runtime": 12.405,
      "eval_samples_per_second": 302.299,
      "eval_steps_per_second": 4.756,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 20.045629501342773,
      "learning_rate": 0.0005,
      "loss": 0.0872,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8981333333333333,
      "eval_loss": 0.4567290246486664,
      "eval_runtime": 12.3885,
      "eval_samples_per_second": 302.7,
      "eval_steps_per_second": 4.762,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}