[Fix] openai: distinguish s3 served-model-name from LoRA adapter syntax (#44450)
Signed-off-by: Muqi Li <muqi1029@gmail.com> Signed-off-by: Xinyuan Tong <xinyuantong.cs@gmail.com> Co-authored-by: Xinyuan Tong <xinyuantong.cs@gmail.com> Co-authored-by: Xinyuan Tong <115166877+JustinTong0323@users.noreply.github.com>
This commit is contained in:
co-authored by
Xinyuan Tong
Xinyuan Tong
parent
2229345dff
commit
16eb87c06f
@@ -40,17 +40,24 @@ class OpenAIServingBase(ABC):
|
||||
def _parse_model_parameter(self, model: str) -> Tuple[str, Optional[str]]:
|
||||
"""Parse 'base-model:adapter-name' syntax to extract LoRA adapter.
|
||||
|
||||
Returns (base_model, adapter_name) or (model, None) if no colon present.
|
||||
Returns (base_model, adapter_name), preserving an exact served model
|
||||
name even when it contains colons (e.g. an object-storage URI).
|
||||
"""
|
||||
if ":" not in model:
|
||||
return model, None
|
||||
|
||||
# Split on first colon only to handle model paths with multiple colons
|
||||
parts = model.split(":", 1)
|
||||
base_model = parts[0].strip()
|
||||
adapter_name = parts[1].strip() or None
|
||||
served_model_name = self.tokenizer_manager.served_model_name
|
||||
if model == served_model_name:
|
||||
return model, None
|
||||
|
||||
return base_model, adapter_name
|
||||
# Match the full served name before splitting URI schemes like s3://.
|
||||
if model.startswith(f"{served_model_name}:"):
|
||||
base_model = served_model_name
|
||||
adapter_name = model[len(served_model_name) + 1 :]
|
||||
else:
|
||||
base_model, adapter_name = model.split(":", 1)
|
||||
|
||||
return base_model.strip(), adapter_name.strip() or None
|
||||
|
||||
def _resolve_lora_path(
|
||||
self,
|
||||
|
||||
Reference in New Issue
Block a user