Skip to content

GitLab

Commit 33650733 authored by zhuwenwen's avatar zhuwenwen
Browse files

skip silu_and_mul_quant

parent d2b52805
Show whitespace changes
Inline Side-by-side
Showing with 8 additions and 8 deletions
vllm/compilation/activation_quant_fusion.py
View file @ 33650733
......@@ -26,14 +26,14 @@ FP4_DTYPE = torch.uint8
SILU_MUL_OP = torch.ops._C.silu_and_mul.default
FUSED_OPS: dict[QuantKey, OpOverload] = {
kFp8StaticTensorSym: torch.ops._C.silu_and_mul_quant.default, # noqa: E501
}
silu_and_mul_nvfp4_quant_supported = (current_platform.is_cuda() and hasattr(
torch.ops._C, "silu_and_mul_nvfp4_quant"))
if silu_and_mul_nvfp4_quant_supported:
FUSED_OPS[
kNvfp4Quant] = torch.ops._C.silu_and_mul_nvfp4_quant.default # noqa: E501
# FUSED_OPS: dict[QuantKey, OpOverload] = {
# kFp8StaticTensorSym: torch.ops._C.silu_and_mul_quant.default, # noqa: E501
# }
# silu_and_mul_nvfp4_quant_supported = (current_platform.is_cuda() and hasattr(
# torch.ops._C, "silu_and_mul_nvfp4_quant"))
# if silu_and_mul_nvfp4_quant_supported:
# FUSED_OPS[
# kNvfp4Quant] = torch.ops._C.silu_and_mul_nvfp4_quant.default # noqa: E501
class ActivationQuantPattern(ABC):
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment