xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-17 16:46:38 +08:00
Code Issues Packages Projects Releases Wiki Activity

[Bugfix][Frontend] Disable embedding API for chat models (#7504)

Browse Source 
Co-authored-by: jack <jack@alex>
This commit is contained in:
jack 2024-08-15 00:15:19 +08:00 committed by GitHub
parent d3d9cb6e4b
commit 67d115db08
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 4 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
vllm/entrypoints/openai/serving_embedding.py
View File

@ -71,7 +71,7 @@ class OpenAIServingEmbedding(OpenAIServing):
lora_modules=None, lora_modules=None,
prompt_adapters=None, prompt_adapters=None,
request_logger=request_logger) request_logger=request_logger)
self._check_embedding_mode(model_config.embedding_mode) self._enabled = self._check_embedding_mode(model_config.embedding_mode)
async def create_embedding( async def create_embedding(
self, self,
@ -83,6 +83,8 @@ class OpenAIServingEmbedding(OpenAIServing):
See https://platform.openai.com/docs/api-reference/embeddings/create See https://platform.openai.com/docs/api-reference/embeddings/create
for the API specification. This API mimics the OpenAI Embedding API. for the API specification. This API mimics the OpenAI Embedding API.
""" """
if not self._enabled:
return self.create_error_response("Embedding API disabled")
error_check_ret = await self._check_model(request) error_check_ret = await self._check_model(request)
if error_check_ret is not None: if error_check_ret is not None:
return error_check_ret return error_check_ret
@ -179,3 +181,4 @@ class OpenAIServingEmbedding(OpenAIServing):
"embedding_mode is False. Embedding API will not work.") "embedding_mode is False. Embedding API will not work.")
else: else:
logger.info("Activating the server engine with embedding enabled.") logger.info("Activating the server engine with embedding enabled.")
return embedding_mode