Commit History

  1. Remove remaining @spaces.GPU decorators from bake_albedo and bake_ao

    68a4e06
  2. Fix ZeroGPU proxy token expiry: single GPU allocation for full pipeline⁠

    5dad3a0
  3. Fix nvdiffrast SM mismatch: add +PTX arch + GL context fallback⁠

    851bece
  4. Preload libcudart.so.13 via ctypes.CDLL before nvdiffrast import

    aead907
  5. Add CUDA 13.0 lib64 to LD_LIBRARY_PATH for nvdiffrast runtime linking

    c901117
  6. Fix nvdiffrast build: set TORCH_CUDA_ARCH_LIST=8.0;9.0 for headless compile

    e0aaa0f
  7. Add full build output to nvdiffrast install for diagnostics

    5167771
  8. Bypass CUDA version mismatch for nvdiffrast build via nvcc shim⁠

    3e8d310
  9. Switch to torch 2.9.1 + runtime nvdiffrast install⁠

    ec19943
  10. Revert to torch 2.6.0 + Jeff Xiang prebuilt nvdiffrast wheel⁠

    be2510d
  11. Install nvdiffrast at runtime instead of build time⁠

    d18c6c9
  12. Fix packages.txt: remove comments (apt-get chokes on them)

    8752554
  13. Fix nvdiffrast ABI mismatch + pymeshfix TrackedArray crash⁠

    09e4a10

  1. Fix pymeshlab pipeline crash + add post-process progress streaming⁠

    f9271f9
  2. Fix all post-processing errors on ZeroGPU/headless environment⁠

    c6d0961
  3. Fix state persistence across ZeroGPU subprocess boundary⁠

    299ba9b
  4. Fix viewer not updating: direct GLB output + allowed_paths⁠

    1fe7d3d
  5. Fix 3D viewer not updating after generation⁠

    00132f3
  6. Add streaming progress to generate_hunyuan + fix omegaconf dep⁠

    bb34c64
  7. Add onnxruntime for rembg background removal

    a525787
  8. Implement Hunyuan3D-2.1 generation with background removal⁠

    0062a30
  9. Fix: preprocess image on remote TRELLIS.2 Space before calling image_to_3d⁠

    8138769
  10. Polish: preset loading, quota recording, coacd fallback⁠

    85a0e9b
  11. Fix: replace PyPI nvdiffrast (doesn't exist) with Jeff Xiang's prebuilt wheel⁠

    eb018ed
  12. Milestones 9-12: UniRig, UE5 Export, Presets, Polish⁠

    75e8e33
  13. Milestones 6+8: albedo/material/AO baking + ORM pack, LODs, collision, pivot, scale⁠

    4537d39
  14. Milestone 4+5: UV unwrap (xatlas), symmetry, and normal map baking (nvdiffrast)⁠

    4fab588
  15. Milestone 3: Stage 2A-2C Mesh Cleanup (repair, cleanup, decimation)⁠

    5c6c726
  16. Switch Stage 1 to gradio_client: all TRELLIS.2 custom wheels ABI-broken on torch 2.8+⁠

    0b410ef
  17. Patch trellis2/representations/mesh/base.py: guard cumesh and flex_gemm imports⁠

    444ea3a
  18. Fix: un-ignore TRELLIS.2/trellis2/models/ — was blocked by 'models/' in .gitignore⁠

    854c275
  19. Fix: evict pip-installed trellis2 stub from sys.modules before loading vendored copy⁠

    5693ff7
  20. Fix: make trellis2/__init__.py fully lazy to survive partial-module namespace⁠

    313ffe0
  21. Fix circular import: bootstrap trellis2 at app.py top-level before src.stages import⁠

    85d4c34
  22. Fix: bump torch to 2.8.0 for ZeroGPU compatibility (2.6.0 no longer supported)

    c00a18a
  23. Milestone 2: Wire TRELLIS.2 generation + vendor both model packages⁠

    1ae114b
  24. Upload app.py

    Verified signature
    56ebf77
  25. Delete app.py

    Verified signature
    87d0a58
  26. Update requirements.txt

    Verified signature
    942c27d
  27. Update README.md

    Verified signature
    8fdae89
  28. Upload README.md

    Verified signature
    8cbbb40
  29. Delete README.md

    Verified signature
    9634403
  30. Upload 2 files

    Verified signature
    3c84e1c
  31. Delete app.py

    Verified signature
    9ef2edf
  32. Delete requirements.txt

    Verified signature
    8e73801
  33. Upload app.py

    Verified signature
    bab10f1
  34. Delete app.py

    Verified signature
    ed96ab6
  35. Upload app.py

    Verified signature
    f0b58dd
  36. Delete app.py

    Verified signature
    883cd00
  37. Create presets/.gitkeep

    Verified signature
    70e2445