reward_config.json

{
  "format_version": 1,
  "base_model_dir": "Qwen3-8B",
  "lora_dir": "step_2200",
  "projector_path": "projector.pth",
  "proj_arch": "mlp2x_relu",
  "torch_dtype": "bfloat16",
  "attn_implementation": null,
  "trust_remote_code": true,
  "reward_activation": "sigmoid"
}