Skip to content

[torch.compile] Turn on silu+fp4 quant fusion by default for O1+ #34718

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
ProExpertProg merged 2 commits into from
Feb 18, 2026
+10 −5
Merged

[torch.compile] Turn on silu+fp4 quant fusion by default for O1+ #34718

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
238f271
Turn on silu+fp4 quant fusion by default
ProExpertProg Feb 17, 2026
2e55859
Fix model_config=None
ProExpertProg Feb 17, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 10 additions & 5 deletions vllm/config/vllm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -95,11 +95,16 @@ def enable_norm_fusion(cfg: "VllmConfig") -> bool:


def enable_act_fusion(cfg: "VllmConfig") -> bool:
"""Enable if either SiLU+Mul or quant FP8 custom op is active;
otherwise Inductor handles fusion."""
return cfg.compilation_config.is_custom_op_enabled(
"silu_and_mul"
) or cfg.compilation_config.is_custom_op_enabled("quant_fp8")
"""
Enable if either SiLU+Mul or quant FP8 custom op is active;
otherwise Inductor handles fusion.
Also enable for FP4 models as FP4 quant is always custom so Inductor cannot fuse it.
"""
return (
cfg.compilation_config.is_custom_op_enabled("silu_and_mul")
or cfg.compilation_config.is_custom_op_enabled("quant_fp8")
or (cfg.model_config is not None and cfg.model_config.is_nvfp4_quantized())
)


def enable_allreduce_rms_fusion(cfg: "VllmConfig") -> bool:
Expand Down