Aryan Kumar 7203b4976e [bugfix]: compile only the LoRA delta, not the wrapped base layer ...

BaseLayerWithLoRA.forward was decorated with torch.compile, so wrapping a linear also compiled its base layer.
Compiling a packed NVFP4/FP8 linear on its own changes its numerics, so a wrapped layer at strength 0 could differ
from the unwrapped layer. forward now adds no compiled region: the base layer runs exactly as it would unwrapped
(eagerly, or traced with a caller's compiled block), and only the adapter delta is compiled.

2026-10-28 14:49:15 -07:00

..

linear.py

[bugfix]: compile only the LoRA delta, not the wrapped base layer

2026-10-28 14:49:15 -07:00