trainer_state.json

{
  "best_metric": 0.8976,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0283/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 35.66513442993164,
      "learning_rate": 0.0002666666666666666,
      "loss": 1.5308,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8144,
      "eval_loss": 0.5837020874023438,
      "eval_runtime": 15.4861,
      "eval_samples_per_second": 242.153,
      "eval_steps_per_second": 3.81,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 38.267066955566406,
      "learning_rate": 0.0002333333333333333,
      "loss": 0.4427,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8648,
      "eval_loss": 0.42420676350593567,
      "eval_runtime": 15.524,
      "eval_samples_per_second": 241.562,
      "eval_steps_per_second": 3.801,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 43.720558166503906,
      "learning_rate": 0.00019999999999999998,
      "loss": 0.2708,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8786666666666667,
      "eval_loss": 0.4109547436237335,
      "eval_runtime": 15.6431,
      "eval_samples_per_second": 239.722,
      "eval_steps_per_second": 3.772,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 42.3889045715332,
      "learning_rate": 0.00016666666666666666,
      "loss": 0.1693,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8869333333333334,
      "eval_loss": 0.3805542290210724,
      "eval_runtime": 15.2471,
      "eval_samples_per_second": 245.949,
      "eval_steps_per_second": 3.87,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 41.54035949707031,
      "learning_rate": 0.0001333333333333333,
      "loss": 0.1089,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8858666666666667,
      "eval_loss": 0.4103439152240753,
      "eval_runtime": 15.9796,
      "eval_samples_per_second": 234.674,
      "eval_steps_per_second": 3.692,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 58.16655349731445,
      "learning_rate": 9.999999999999999e-05,
      "loss": 0.076,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8952,
      "eval_loss": 0.39012566208839417,
      "eval_runtime": 16.6213,
      "eval_samples_per_second": 225.614,
      "eval_steps_per_second": 3.55,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 120.32537841796875,
      "learning_rate": 6.666666666666666e-05,
      "loss": 0.0511,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8906666666666667,
      "eval_loss": 0.416597843170166,
      "eval_runtime": 15.587,
      "eval_samples_per_second": 240.586,
      "eval_steps_per_second": 3.785,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 105.58467864990234,
      "learning_rate": 3.333333333333333e-05,
      "loss": 0.0412,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8976,
      "eval_loss": 0.4149155020713806,
      "eval_runtime": 15.4333,
      "eval_samples_per_second": 242.981,
      "eval_steps_per_second": 3.823,
      "step": 2664
    }
  ],
  "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": 6.5411433916416e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}