Repository navigation
Test the remote backends against llama-server - #426
Merged
Merged
Conversation
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This PR adds an integration suite that runs
OpenAILanguageModel(Chat Completions and Responses),OpenResponsesLanguageModel, andAnthropicLanguageModelagainst a local llama-server, the HTTP server in llama.cpp. These backends take a base URL, so they work with it, but nothing showed that, and the README didn't mention it. The suite runs only whenLLAMA_SERVER_BASE_URLis set, and its image tests only whenLLAMA_SERVER_VISION_BASE_URLis set. It covers text and streamed responses with usage, instructions, multi-turn sessions, structured output, tool calling, reasoning, image input, temperature, maximum tokens, stop sequences, cancellation, and a prompt longer than the context.Where llama-server differs from the API it imitates, the test marks the difference as a known issue instead of working around it. With tools and a response schema in one request, the Chat Completions and Responses endpoints allow only the schema, so the model can't call a tool. The Messages endpoint ignores
output_config.format, and it has no way to turn off parallel tool calls. The README gets a llama-server section with an example and these limits.