da3metric-large Model appears NOT to support pose-conditioned inference (cam_enc is None).

Hi, thank you very much for your open-source work.

I'm using the

> da3metric-large

model and trying to use the **pose-conditioned method**, but it's throwing an exception.

**this is my code:**
> # Run inference (pose-conditioned if model supports it)
> try:
> prediction = model.inference(
> image=image_list,
> extrinsics=extrinsics_array,
> intrinsics=intrinsics_array,
> align_to_input_ext_scale=args.align_to_input_ext_scale,
> infer_gs=args.infer_gs,
> use_ray_pose=args.use_ray_pose,
> process_res=args.process_res,
> process_res_method=args.process_res_method,
> export_dir=str(args.export_dir),
> export_format=args.export_format,
> )
> except TypeError as e:
> # Typical case: cam_enc is None -> NoneType not callable
> msg = str(e)
> if "NoneType" in msg and "callable" in msg:
> print("[WARN] Model appears NOT to support pose-conditioned inference (cam_enc is None).")
> print("[WARN] Fallback: run inference WITHOUT intrinsics/extrinsics; still export npz.")
> prediction = model.inference(
> image=image_list,
> process_res=args.process_res,
> process_res_method=args.process_res_method,
> export_dir=str(args.export_dir),
> export_format=args.export_format,
> )
> else:
> raise

**cam_enc is None -> NoneType not callable log:**
> [INFO] Using 6 images.
> [INFO] intrinsics_array: (6, 3, 3) float32
> [INFO] extrinsics_array: (6, 4, 4) float32
> [INFO ] using MLP layer as FFN
> [INFO ] Processed Images Done taking 0.44652867317199707 seconds. Shape: torch.Size([6, 3, 2044, 2044])
> [WARN] **Model appears NOT to support pose-conditioned inference (cam_enc is None)**.
> [WARN] **Fallback: run inference WITHOUT intrinsics/extrinsics; still export npz.**
>

Does the Metric model not support pose-conditioned inference?
The Metric model supports preset intrinsic parameters, but does not support inference with extrinsic parameters.