Where to set the layout_post_optimization_method parameter for the run() method?

I'm using the run() method of the SAM 3D model (specifically the core inference interface) and noticed that it includes a parameter self.layout_post_optimization_method which controls the layout post-optimization process.

According to the code logic:
When with_layout_postprocess=True (default), the method attempts to run layout post-optimization using self.layout_post_optimization_method.
However, I cannot find any documentation, configuration options, or setter methods in the project to specify/modify this parameter.

def run(
self,
image: Union[None, Image.Image, np.ndarray],
mask: Union[None, Image.Image, np.ndarray] = None,
seed: Optional[int] = None,
stage1_only=False,
with_mesh_postprocess=True,
with_texture_baking=True,
with_layout_postprocess=True,
use_vertex_color=False,
stage1_inference_steps=None,
stage2_inference_steps=None,
use_stage1_distillation=False,
use_stage2_distillation=False,
pointmap=None,
decode_formats=None,
estimate_plane=False,
) -> dict:
#........

try:
if (
with_layout_postprocess
and self.layout_post_optimization_method is not None
):
assert glb is not None, "require mesh to run postprocessing"
logger.info("Running layout post optimization method...")
postprocessed_pose = self.run_post_optimization(
deepcopy(glb),
pointmap_dict["intrinsics"],
ss_return_dict,
ss_input_dict,
)
ss_return_dict.update(postprocessed_pose)
except Exception as e:
logger.error(
f"Error during layout post optimization: {e}", exc_info=True
)

# glb.export("sample.glb")
logger.info("Finished!")

return {
**ss_return_dict,
**outputs,
"pointmap": pts.cpu().permute((1, 2, 0)), # HxWx3
"pointmap_colors": pts_colors.cpu().permute((1, 2, 0)), # HxWx3
}