fix: support Ollama 'reasoning' field alongside 'reasoning_content'#792
Open
Br1an67 wants to merge 1 commit intoQwenLM:mainfrom
Open
fix: support Ollama 'reasoning' field alongside 'reasoning_content'#792Br1an67 wants to merge 1 commit intoQwenLM:mainfrom
Br1an67 wants to merge 1 commit intoQwenLM:mainfrom
Conversation
…ming Ollama returns thinking content in a 'reasoning' field on the streaming chunk delta, while DashScope/DeepSeek-style APIs use 'reasoning_content'. The current code only checks 'reasoning_content', causing thinking content from Ollama to be silently discarded. Use getattr with fallback to check both field names in all three code paths (delta_stream, non-delta stream, and non-stream). Fixes QwenLM#789
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Problem
When using Qwen3 models via Ollama, thinking content arrives in a
reasoningfield on the streaming chunk delta. Howeveroai.pyonly checks forreasoning_content, so thinking content from Ollama is silently discarded andfull_reasoning_contentis never populated.This also breaks the agentic loop when
thought_in_content=True, because the postprocessor waits for</think>tags that never appear.Fix
Use
getattrwith fallback to check bothreasoning_content(DashScope/DeepSeek) andreasoning(Ollama) field names in all three code paths:Ollama chunk format
vs DashScope/DeepSeek:
Fixes #789