trainer_state.json

{
  "best_metric": 0.8890666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0818/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": 384.4679870605469,
      "learning_rate": 9.504844339512095e-05,
      "loss": 2.9062,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7405333333333334,
      "eval_loss": 1.05662202835083,
      "eval_runtime": 13.596,
      "eval_samples_per_second": 275.817,
      "eval_steps_per_second": 4.34,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 95.58940887451172,
      "learning_rate": 8.117449009293668e-05,
      "loss": 0.7375,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8469333333333333,
      "eval_loss": 0.5305494666099548,
      "eval_runtime": 13.3789,
      "eval_samples_per_second": 280.292,
      "eval_steps_per_second": 4.41,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 73.2638168334961,
      "learning_rate": 6.112604669781572e-05,
      "loss": 0.46,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8757333333333334,
      "eval_loss": 0.42302221059799194,
      "eval_runtime": 13.5543,
      "eval_samples_per_second": 276.665,
      "eval_steps_per_second": 4.353,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 83.4851303100586,
      "learning_rate": 3.887395330218429e-05,
      "loss": 0.3555,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.876,
      "eval_loss": 0.411223828792572,
      "eval_runtime": 13.6226,
      "eval_samples_per_second": 275.277,
      "eval_steps_per_second": 4.331,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 75.92252349853516,
      "learning_rate": 1.8825509907063327e-05,
      "loss": 0.2855,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8888,
      "eval_loss": 0.38055911660194397,
      "eval_runtime": 13.4524,
      "eval_samples_per_second": 278.761,
      "eval_steps_per_second": 4.386,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 90.3882827758789,
      "learning_rate": 4.951556604879048e-06,
      "loss": 0.2576,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8864,
      "eval_loss": 0.3739990293979645,
      "eval_runtime": 13.5078,
      "eval_samples_per_second": 277.618,
      "eval_steps_per_second": 4.368,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 120.70484924316406,
      "learning_rate": 0.0,
      "loss": 0.2387,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8890666666666667,
      "eval_loss": 0.3630889356136322,
      "eval_runtime": 13.4688,
      "eval_samples_per_second": 278.421,
      "eval_steps_per_second": 4.38,
      "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
}