fix: bypass LiteLLM for Ollama embeddings to resolve 400 Bad Request (v1.9 regression)#1509
Open
GratefulDave wants to merge 4 commits intoagent0ai:mainfrom
Open
fix: bypass LiteLLM for Ollama embeddings to resolve 400 Bad Request (v1.9 regression)#1509GratefulDave wants to merge 4 commits intoagent0ai:mainfrom
GratefulDave wants to merge 4 commits intoagent0ai:mainfrom
Conversation
LiteLLM's Ollama embedding handler sends a malformed request to Ollama's /api/embed endpoint, causing a 400 Bad Request error on Ollama 0.18.x. - Add `_ollama_embed()` to `LiteLLMEmbeddingWrapper` that calls Ollama's `/api/embed` directly via httpx, stripping the "ollama/" prefix from the model name (the root cause of the malformed request) - Route `embed_query` and `embed_documents` through this helper when provider == "ollama", bypassing LiteLLM entirely - Wrap `search_similarity_threshold` in try/except so an embedding failure returns [] instead of crashing the agent Fixes agent0ai#1425 Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
After the v1.9 refactor of
LiteLLMEmbeddingWrapper, the_ollama_embed()bypass that was previously working was removed. This reintroduced the Ollama embedding 400 Bad Request error for users running Ollama as their embedding provider.Error:
Root Cause
LiteLLM's Ollama embedding handler sends a malformed request to Ollama's
/api/embedendpoint on Ollama 0.18.x+. Specifically, it passes the model name with theollama/prefix intact (e.g.ollama/nomic-embed-text) which Ollama does not recognize.Fix
_ollama_embed()toLiteLLMEmbeddingWrapper— calls Ollama's/api/embeddirectly viahttpx, stripping theollama/prefix from the model nameembed_queryandembed_documentsthrough this helper whenprovider == "ollama", bypassing LiteLLM entirelysearch_similarity_thresholdin try/except so an embedding failure returns[]instead of crashing the agentTesting
Verified working against Ollama 0.18.x with
nomic-embed-textmodel overhost.docker.internal:11434.Fixes #1425