EngineX-Hygon/sglang
5
0
Fork 0
Code Issues Pull Requests Actions 7 Projects Releases Wiki Activity

fix: update model runner (#5934)

Browse Source
This commit is contained in:
Yineng Zhang
2025-04-30 19:49:26 -07:00
committed by GitHub
parent 256c4c2519
commit 8441baad6e
1 changed files with 1 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
python/sglang/srt/model_executor/model_runner.py
View File

@@ -81,7 +81,6 @@ from sglang.srt.utils import (
get_available_gpu_memory, get_available_gpu_memory,
get_bool_env_var, get_bool_env_var,
init_custom_process_group, init_custom_process_group,
is_ampere_with_cuda_12_3,
is_cuda, is_cuda,
is_fa3_default_architecture, is_fa3_default_architecture,
is_flashinfer_available, is_flashinfer_available,
@@ -264,7 +263,7 @@ class ModelRunner:
if not self.use_mla_backend: if not self.use_mla_backend:
# MHA architecture # MHA architecture
if ( if (
(is_ampere_with_cuda_12_3() or is_hopper_with_cuda_12_3()) is_hopper_with_cuda_12_3()
and is_no_spec_infer_or_topk_one(server_args) and is_no_spec_infer_or_topk_one(server_args)
and is_fa3_default_architecture(self.model_config.hf_config) and is_fa3_default_architecture(self.model_config.hf_config)
): ):