fix(zerogpu): lazy-import nvdiffrast in postprocessing_utils

`import nvdiffrast.torch as dr` at module top level triggered CUDA init
in the main process via the import chain:
app.py
→ trellis.pipelines.NeARImageToRelightable3DPipeline
→ trellis.utils.render_utils_rl
→ trellis.utils.postprocessing_utils
→ import nvdiffrast.torch as dr ← CUDA init here

This caused ZeroGPU's stateless-GPU enforcement to fire even before any
@spaces.GPU callback ran, blocking every GPU callback silently.

Fix: remove top-level import; add `import nvdiffrast.torch as dr` lazily
inside the `mode == 'opt'` branch where it is actually used.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

trellis/utils/postprocessing_utils.py

@@ -2,7 +2,6 @@ from typing import *
 import numpy as np
 import torch
 import utils3d
-import nvdiffrast.torch as dr
 from tqdm import tqdm
 import trimesh
 import trimesh.visual
@@ -346,6 +345,7 @@ def bake_texture(
         texture = cv2.inpaint(texture, mask, 3, cv2.INPAINT_TELEA)
 
     elif mode == 'opt':
+        import nvdiffrast.torch as dr  # lazy: needs active CUDA context
         rastctx = utils3d.torch.RastContext(backend='cuda')
         observations = [observations.flip(0) for observations in observations]
         masks = [m.flip(0) for m in masks]