{
  "pretrained_model": "./experiments/baseline/checkpoint/latest.pth",
  "args_file": "./experiments/baseline/checkpoint/args.json",
  "train_set": "./preprocessed_data",
  "train_set_list": "/home/Humble/Humble/EndtoEnd/YSH-Diffusion-Planner/diffusion_planner_training.json",
  "num_samples": 8,
  "beta": 1.0,
  "weight_clip": 10.0,
  "temperature": 1.0,
  "learning_rate": 1e-05,
  "gradient_accumulation_steps": 1,
  "reward_type": "base",
  "epochs": 50,
  "batch_size": 256,
  "num_workers": 16,
  "save_freq": 5,
  "output_dir": "./wr_output",
  "name": "wr-base",
  "use_mlflow": true,
  "mlflow_tracking_uri": "https://dagshub.com/YSH-research/YSH-Diffusion-Planner.mlflow",
  "mlflow_experiment_name": "wr-finetuning",
  "local_rank": -1
}