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

feat: add add_chunked_prefix_cache_attention_backend (#11636)

Browse Source
This commit is contained in:
Yineng Zhang
2025-10-14 21:48:13 -07:00
committed by GitHub
parent 780fbf2f38
commit 91fc5bb5a9
1 changed files with 8 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
python/sglang/srt/model_executor/model_runner.py
View File

@@ -190,6 +190,14 @@ def add_mla_attention_backend(backend_name):
logger.info(f"Added {backend_name} to MLA_ATTENTION_BACKENDS.")
def add_chunked_prefix_cache_attention_backend(backend_name):
if backend_name not in CHUNKED_PREFIX_CACHE_SUPPORTED_ATTENTION_BACKENDS:
CHUNKED_PREFIX_CACHE_SUPPORTED_ATTENTION_BACKENDS.append(backend_name)
logger.info(
f"Added {backend_name} to CHUNKED_PREFIX_CACHE_SUPPORTED_ATTENTION_BACKENDS."
)
_is_hip = is_hip()
_is_npu = is_npu()
_is_cpu_amx_available = cpu_has_amx_support()