Repository navigation
fix(llm): providers implement abstract _chat_completion_impl — Ollama/all chat was dead (#11249) - #11250
Merged
Conversation
…erriding the wrapper) — Ollama/chat was dead BaseProvider.chat_completion is a concrete wrapper (rate-limiter + backoff) around the abstract _chat_completion_impl. Five providers overrode the wrapper instead of implementing the abstract, leaving __abstractmethods__ non-empty so the class could not instantiate — registration threw (caught at debug) and the provider silently vanished. With Ollama gone and Bedrock unconfigured, the registry had no available provider and ALL chat failed (Company OS CEO chat returned only the clarify fallback). Rename the override to _chat_completion_impl in OllamaProvider, CustomOpenAI, Groq, HuggingFace, VLLMBase; they now also run inside the base rate-limit/backoff wrapper. Verified all five __abstractmethods__ empty; Ollama registers + chats live.
Contributor
✅ SSOT Configuration Compliance: Passing🎉 No hardcoded values detected that have SSOT config equivalents! |
Owner
Author
|
Heads-up (parallel-session overlap): #11249 was resolved by #11257 (merged to Dev_new_gui as b58eb0c) — same fix (rename |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Thinking Path
User: "Company OS not loading" + "CEO chat just responds with a single phrase — Could you clarify? intent". Two bugs: (1) Company OS 500s from an unapplied DB migration (fixed separately on the box — see below), and (2) CEO chat → traced to
provider_registry: No available provider for chat request. Ollama is up with models but the registry only had['bedrock']. Root cause:OllamaProvider(and 4 others) override the concreteBaseProvider.chat_completionwrapper instead of implementing the abstract_chat_completion_impl, so__abstractmethods__is non-empty →TypeErroron instantiate → registration silently fails → no providers → all chat dead.What Changed
Renamed the
chat_completionoverride →_chat_completion_implin the 5 genuinely-brokenBaseProvidersubclasses:OllamaProvider,CustomOpenAIProvider,GroqProvider,HuggingFaceProvider,VLLMBaseProvider. They now also correctly run inside the base rate-limiter/backoff wrapper. (ollama.py/vllm.py/transformers/mock are standalone classes, not affected.)Verification
py_compileclean on all 5.__abstractmethods__==[](instantiable).['ollama','bedrock'];svc.chat()returns real content via ollama/qwen3.5:9b (was empty). CEO chat restored.Model Used
Opus 4.8 (1M context)
Closes #11249