Add env var to control custom Triton kernel cache and set CSGMV as default backend. (#12176)
This commit is contained in:
@@ -511,7 +511,7 @@ class SRTRunner:
|
||||
attention_backend: Optional[str] = None,
|
||||
prefill_attention_backend: Optional[str] = None,
|
||||
decode_attention_backend: Optional[str] = None,
|
||||
lora_backend: str = "triton",
|
||||
lora_backend: str = "csgmv",
|
||||
disable_cuda_graph: bool = False,
|
||||
disable_radix_cache: bool = False,
|
||||
chunked_prefill_size: Optional[int] = None,
|
||||
|
||||
Reference in New Issue
Block a user