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 89

fix(deepseek): guard reasoning_content injection to thinking-mode conversations only

Browse source
This commit is contained in:
Ishaan Jaffer 2026-04-27 18:49:02 -07:00
parent 3401e0aa54
commit b3d31d01eb
No known key found for this signature in database
1 changed files with 14 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

18
litellm/llms/deepseek/chat/transformation.py
View file

@ -80,12 +80,22 @@ class DeepSeekChatConfig(OpenAIGPTConfig):
) -> List[AllMessageValues]:
"""
DeepSeek V4 thinking mode requires `reasoning_content` on every assistant message
in the conversation history. If an assistant message is missing it (e.g. because
Message.__init__ deleted it when it was None), inject an empty string so the API
does not return 400 "reasoning_content must be passed back".
in the conversation history. Only inject when we can confirm this is a thinking-mode
conversation: at least one assistant message already carries `reasoning_content`
(returned by DeepSeek on a prior turn). If no message has it the request is not
in thinking mode and we leave payloads untouched.
"""
thinking_active = any(
msg.get("role") == "assistant" and "reasoning_content" in msg
for msg in messages
)
if not thinking_active:
return messages
for message in messages:
if message.get("role") == "assistant" and "reasoning_content" not in message:
if (
message.get("role") == "assistant"
and "reasoning_content" not in message
):
message["reasoning_content"] = "" # type: ignore[typeddict-unknown-key]
return messages