@tduncan / Nvidia-Optimization-Stack-VLLM Mirror · read-only / commits / ba8dc958a3
Fork 0

A high-throughput and memory-efficient inference and serving engine for LLMs

Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git

Synced Nov 17, 04:15 UTC

[Minor] Fix bias in if to remove ambiguity (#3259)

Hui Liu committed Mar 13, 2024 at 09:16 UTC ba8dc958a3d8533a6e5b7debda47e4d42a062b78
1 file changed +1 -1
vllm/model_executor/layers/linear.py
+1 -1
@@ -73,7 +73,7 @@ class UnquantizedLinearMethod(LinearMethodBase):
73 bias: Optional[torch.Tensor] = None) -> torch.Tensor:
74 weight = weights["weight"]
75 if self.separate_bias_add:
76 - if bias:
76 + if bias is not None:
77 return F.linear(x, weight) + bias
78 return F.linear(x, weight)
79 return F.linear(x, weight, bias)