scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-09-12 23:01:41 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 87

fix(proxy): - Fixed issue where Volcengine embedding models were incorrectly identified as "text-embedding-3-small"

Browse source 
- Implemented route type-based model selection in proxy server
- Use embedding_model setting for embedding requests instead of completion_model
- Ensure user-provided model names take precedence over default settings
- Maintain compatibility with completion requests using completion_model setting
This commit is contained in:
huanghaiyang 2026-04-08 14:43:55 +08:00
parent 03d9746815
commit 072214c7bb
1 changed files with 15 additions and 6 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

21
litellm/proxy/common_request_processing.py
View file

@ -700,12 +700,21 @@ class ProxyBaseLLMRequestProcessing:
"queue_time_seconds"
] = queue_time_seconds
self.data["model"] = (
general_settings.get("completion_model", None) # server default
or user_model # model name passed via cli args
or model # for azure deployments
or self.data.get("model", None) # default passed in http request
)
# Use appropriate model setting based on route type
if route_type == "aembedding":
self.data["model"] = (
general_settings.get("embedding_model", None) # server default for embedding
or user_model # model name passed via cli args
or model # for azure deployments
or self.data.get("model", None) # default passed in http request
)
else:
self.data["model"] = (
general_settings.get("completion_model", None) # server default for completion
or user_model # model name passed via cli args
or model # for azure deployments
or self.data.get("model", None) # default passed in http request
)
# override with user settings, these are params passed via cli
if user_temperature: