minke.yu/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 13 Packages Projects Releases Wiki Activity

[HiCache] fix condition check when use decode offload (#14489)

Browse Source
This commit is contained in:
Francis
2025-12-08 15:52:40 -08:00
committed by GitHub
parent 93043f7b13
commit 07404d7689
1 changed files with 4 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
python/sglang/srt/server_args.py
+4 -1
View File
@@ -1668,7 +1668,10 @@ class ServerArgs:
"Page first direct layout only support direct io backend" "Page first direct layout only support direct io backend"
) )
if self.enable_hierarchical_cache and self.hicache_io_backend == "kernel": if (
self.enable_hierarchical_cache
or self.disaggregation_decode_enable_offload_kvcache
) and self.hicache_io_backend == "kernel":
# fix for the compatibility issue with FlashAttention3 decoding and HiCache kernel backend # fix for the compatibility issue with FlashAttention3 decoding and HiCache kernel backend
if self.decode_attention_backend is None: if self.decode_attention_backend is None:
if not self.use_mla_backend(): if not self.use_mla_backend():