fix: update package_cuda.py for PyInstaller 6.18 layout and remove split_binary.py

- Fix is_nvidia_file() to match NVIDIA DLLs in _internal/torch/lib/
  (PyInstaller 6.18 + torch 2.10 no longer uses nvidia/ subdirectories)
- Remove deprecated split_binary.py (both archives are under 2GB)
- Update torch_compat range to >=2.6.0,<2.11.0
- Update build docs for new dual-archive packaging flow
This commit is contained in:
Jamie Pine
2026-03-17 04:57:05 -07:00
parent 564d787927
commit 88be097b62
5 changed files with 54 additions and 114 deletions
+1 -1
View File
@@ -170,4 +170,4 @@ PyInstaller onedir creates a parent bootloader + child Python process on Windows
5. Update `main.rs`: change launch path to `backends/cuda/` dir + add `.current_dir()`
6. Add `ensure_cuda_structure()` helper in Rust to verify exe + nvidia/ subdirs exist before spawning
7. Update CI pipeline: `build-cuda-windows` produces two archives instead of split parts
8. Update `split_binary.py` or replace with archive-based distribution
8. ~~Update `split_binary.py` or replace with archive-based distribution~~ Done: replaced with `package_cuda.py`