trainer_state.json

{
  "best_metric": 0.9016,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0561/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": 99.12455749511719,
      "learning_rate": 8e-05,
      "loss": 2.8825,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7421333333333333,
      "eval_loss": 1.1071349382400513,
      "eval_runtime": 14.3808,
      "eval_samples_per_second": 260.764,
      "eval_steps_per_second": 4.103,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 145.2089385986328,
      "learning_rate": 7.000000000000001e-05,
      "loss": 0.7239,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8493333333333334,
      "eval_loss": 0.5343993306159973,
      "eval_runtime": 14.2363,
      "eval_samples_per_second": 263.411,
      "eval_steps_per_second": 4.144,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 154.9996337890625,
      "learning_rate": 6e-05,
      "loss": 0.4122,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8776,
      "eval_loss": 0.4187747836112976,
      "eval_runtime": 14.2192,
      "eval_samples_per_second": 263.728,
      "eval_steps_per_second": 4.149,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 204.33731079101562,
      "learning_rate": 5e-05,
      "loss": 0.2776,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8802666666666666,
      "eval_loss": 0.3913143575191498,
      "eval_runtime": 13.8607,
      "eval_samples_per_second": 270.549,
      "eval_steps_per_second": 4.257,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 131.89024353027344,
      "learning_rate": 4e-05,
      "loss": 0.1944,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8936,
      "eval_loss": 0.36734721064567566,
      "eval_runtime": 14.1479,
      "eval_samples_per_second": 265.057,
      "eval_steps_per_second": 4.17,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 110.04642486572266,
      "learning_rate": 3e-05,
      "loss": 0.1446,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8957333333333334,
      "eval_loss": 0.3669804334640503,
      "eval_runtime": 14.3243,
      "eval_samples_per_second": 261.793,
      "eval_steps_per_second": 4.119,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 258.53656005859375,
      "learning_rate": 2e-05,
      "loss": 0.1113,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9016,
      "eval_loss": 0.361631840467453,
      "eval_runtime": 14.1791,
      "eval_samples_per_second": 264.474,
      "eval_steps_per_second": 4.161,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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
}