Adrien Toupet fb2b6c7d4c Fix BlockSwap not applying on cached models and rename offload_io_components to swap_io_components ...

- Fixed critical bug where BlockSwap configuration changes were not applied to cached models
- Apply BlockSwap immediately in _handle_blockswap_config instead of deferring to materialization phase
- Renamed offload_io_components to swap_io_components across entire codebase for consistency
- Removed unused _pending_blockswap_config attribute

2025-10-10 17:48:12 -04:00

..

common

Fix GGUF implementation: centralize warning suppression, improve type hints, enhance error handling

2025-09-24 15:47:18 -04:00

core

Fix BlockSwap not applying on cached models and rename offload_io_components to swap_io_components

2025-10-10 17:48:12 -04:00

data/image/transforms

Restored compatibility with FP8 safetensors for MPS backend; Changed the way to detect MPS device

2025-08-13 00:53:05 +08:00

interfaces

Fix BlockSwap not applying on cached models and rename offload_io_components to swap_io_components

2025-10-10 17:48:12 -04:00

models

feat: implement torch.compile optimization with BlockSwap compatibility

2025-10-10 12:34:49 -04:00

optimization

Fix BlockSwap not applying on cached models and rename offload_io_components to swap_io_components

2025-10-10 17:48:12 -04:00

utils

feat: implement torch.compile optimization with BlockSwap compatibility

2025-10-10 12:34:49 -04:00

__init__.py

Simplify __init__.py files + fix dependencies + add pyproject.toml

2025-09-22 15:47:04 -04:00