trainer_state.json

{
  "best_metric": 0.7754666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0468/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": 111.7046127319336,
      "learning_rate": 2.8514533018536286e-05,
      "loss": 3.8462,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.37493333333333334,
      "eval_loss": 3.751643657684326,
      "eval_runtime": 15.4505,
      "eval_samples_per_second": 242.71,
      "eval_steps_per_second": 3.819,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 165.11788940429688,
      "learning_rate": 2.4352347027881003e-05,
      "loss": 3.1743,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5546666666666666,
      "eval_loss": 2.5535027980804443,
      "eval_runtime": 15.4215,
      "eval_samples_per_second": 243.168,
      "eval_steps_per_second": 3.826,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 119.3648910522461,
      "learning_rate": 1.8337814009344716e-05,
      "loss": 1.8878,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6808,
      "eval_loss": 1.3730175495147705,
      "eval_runtime": 15.4769,
      "eval_samples_per_second": 242.296,
      "eval_steps_per_second": 3.812,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 296.2908630371094,
      "learning_rate": 1.1662185990655285e-05,
      "loss": 1.2369,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7536,
      "eval_loss": 1.0279537439346313,
      "eval_runtime": 15.2669,
      "eval_samples_per_second": 245.629,
      "eval_steps_per_second": 3.865,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 165.6299591064453,
      "learning_rate": 5.647652972118998e-06,
      "loss": 1.0068,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7669333333333334,
      "eval_loss": 0.9321514368057251,
      "eval_runtime": 15.441,
      "eval_samples_per_second": 242.861,
      "eval_steps_per_second": 3.821,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 247.4042205810547,
      "learning_rate": 1.4854669814637145e-06,
      "loss": 0.9389,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7738666666666667,
      "eval_loss": 0.8681606650352478,
      "eval_runtime": 15.4329,
      "eval_samples_per_second": 242.987,
      "eval_steps_per_second": 3.823,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 97.57371520996094,
      "learning_rate": 0.0,
      "loss": 0.9157,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7754666666666666,
      "eval_loss": 0.8891664743423462,
      "eval_runtime": 15.5629,
      "eval_samples_per_second": 240.957,
      "eval_steps_per_second": 3.791,
      "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
}