MBP M5 32GB -- 524.9s (8m45s)

[Modjular]

(.venv) tony@Tonys-MacBook-Pro-2 trellis-mac % python generate.py /Users/tony/Downloads/Gemini_Generated_Image_3770hs3770hs3770.png
============================================================
TRELLIS.2 on Apple Silicon
============================================================

Loading pipeline...

...cut the downloads for brevity...

Device: MPS
Input: /Users/tony/Downloads/Gemini_Generated_Image_3770hs3770hs3770.png (2816x1536)

Generating 3D model (pipeline=512, seed=42)...
Sampling sparse structure: 100%|██████████████████████████████████████████████████████████████████████████████████████| 12/12 [01:38<00:00,  8.18s/it]
Sampling shape SLat:   0%|                                                                                                     | 0/12 [00:00<?, ?it/s]/Users/tony/Documents/scratch/20260428_trellis2/trellis-mac/TRELLIS.2/trellis2/modules/sparse/basic.py:283: UserWarning: The operator 'aten::segment_reduce' is not currently supported on the MPS backend and will fall back to run on the CPU. This may have performance implications. (Triggered internally at /Users/runner/work/pytorch/pytorch/pytorch/aten/src/ATen/mps/MPSFallback.mm:34.)
  red = torch.segment_reduce(red, reduce=op, lengths=self.seqlen)
Sampling shape SLat: 100%|████████████████████████████████████████████████████████████████████████████████████████████| 12/12 [03:22<00:00, 16.87s/it]
Sampling texture SLat: 100%|██████████████████████████████████████████████████████████████████████████████████████████| 12/12 [01:54<00:00,  9.57s/it]

Mesh: 2,837,337 vertices, 5,954,446 triangles
Generation time: 524.9s

Baking PBR textures via Metal (1024x1024)...
  Simplifying mesh: 5,954,446 -> ~200,000 faces
Original mesh: 163013 vertices, 497599 faces
After filling holes: 163095 vertices, 497857 faces
Building BVH for current mesh...Done
Cleaning mesh...
After inital simplification: 163095 vertices, 497857 faces
After initial cleanup: 448484 vertices, 567899 faces
Simplifying [thres=1.00e-05]: 100%|███████████████████████████████████████████████████████████████████████| 367899/367899 [00:01<00:00, 354729.05it/s]
After final simplification: 218610 vertices, 180690 faces
After final cleanup: 142171 vertices, 186693 faces
Done
Parameterizing new mesh...
Get 23157 clusters after fast clustering
Adding clusters to xatlas: 100%|█████████████████████████████████████████████████████████████████████████████| 23157/23157 [00:00<00:00, 50627.79it/s]
xatlas: Computing charts: 100%|████████████████████████████████████████████████████████████████████████████████████| 100/100 [00:00<00:00, 380.42it/s]
xatlas: Packing charts: 100%|██████████████████████████████████████████████████████████████████████████████████████| 100/100 [00:00<00:00, 238.38it/s]
xatlas: Building output meshes: 100%|█████████████████████████████████████████████████████████████████████████████| 100/100 [00:00<00:00, 9651.40it/s]
Gathering results from xatlas: 100%|█████████████████████████████████████████████████████████████████████████| 23157/23157 [00:00<00:00, 42408.51it/s]
Done
Sampling attributes...Done
Finalizing mesh...Detected transparency (alpha min=0), using BLEND mode
Done
  Saved: output_3d.glb
  Bake time: 19s
Saved: output_3d.obj

Total time: 524.9s generation + baking

No bugs, just wanted to report a successful run on an M5. Great job with the port @shivampkumar

[shivampkumar]

Thank you!
Hope you get some cool objects out of it!