Skip to content

GitLab

"...en/git@developer.sourcefind.cn:renzhc/diffusers_dcu.git" did not exist on "ba0e732eb059b9eb3afe4b643be471d93154d1cc"
Unverified Commit 15c75e41 authored by Lianmin Zheng's avatar Lianmin Zheng Committed by GitHub
Browse files

[Fix] Fix --disable-flashinfer (#1389)

parent 224200e3
Hide whitespace changes
Inline Side-by-side
Showing with 2 additions and 0 deletions
python/sglang/srt/server_args.py
View file @ 15c75e41
...@@ -140,11 +140,13 @@ class ServerArgs: ...@@ -140,11 +140,13 @@ class ServerArgs:
"The option '--disable-flashinfer' will be deprecated in the next release. " "The option '--disable-flashinfer' will be deprecated in the next release. "
"Please use '--attention-backend triton' instead." "Please use '--attention-backend triton' instead."
) )
self.attention_backend = "triton"
if self.disable_flashinfer_sampling: if self.disable_flashinfer_sampling:
logger.warning( logger.warning(
"The option '--disable-flashinfer-sampling' will be deprecated in the next release. " "The option '--disable-flashinfer-sampling' will be deprecated in the next release. "
"Please use '--sampling-backend pytorch' instead. " "Please use '--sampling-backend pytorch' instead. "
) )
self.sampling_backend = "pytorch"
# Model-specific patches # Model-specific patches
if "Alibaba-NLP/gte-Qwen2-1.5B-instruct" == self.model_path: if "Alibaba-NLP/gte-Qwen2-1.5B-instruct" == self.model_path:
......
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