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 9, 14:45 UTC
[Bugfix] Fix Can't instantiate abstract class DeepseekV32IndexerBackend (#33052)
Signed-off-by: chaunceyjiang <chaunceyjiang@gmail.com>
Chauncey committed
Jan 26, 2026 at 22:44 UTC
9ef3b718d98ada5c02ebe67c2a545d3005a6b10d
1 file changed
+1
-1
vllm/distributed/kv_transfer/kv_connector/v1/nixl_connector.py
+1
-1
@@ -302,7 +302,7 @@ class NixlConnector(KVConnectorBase_V1):
302
@property
303
def prefer_cross_layer_blocks(self) -> bool:
304
backend = get_current_attn_backend(self._vllm_config)
305
- if backend().get_name() not in (
305
+ if backend.get_name() not in (
306
"FLASH_ATTN",
307
"FLASHINFER",
308
):