Repository navigation
Benchmark Muse Spark Open and GPT-6.1 Terra - #68
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Publishes complete benchmarks for Muse Spark Open and GPT-6.1 Terra: 76 full result files and 10,092 samples. Each model has three runs at Low/High for the six original tasks and one run at each effort for Text. Includes regenerated quality, cost and speed leaderboards and the website summary.
Muse uses OpenRouter/Together with no-training filtering and named XYXY 0–1000 coordinates selected from three 50-image format comparisons. Terra uses the native OpenAI Responses API and the existing OpenAI absolute uploaded-image coordinate format; its additional format search was explicitly waived. The split preserves every result byte and score; no inference was rerun.
341 tests pass, Ruff check/format pass, protocol validation passes with 266 existing historical warnings, and summary freshness is checked. Historical result files are unchanged.
This PR targets
mainand can merge independently. The other four models and campaign tooling remain in follow-up #67. Related Inference integrations: roboflow/inference#3283 and roboflow/inference#3284.Format/provider sources: Muse model endpoint, Meta model card, OpenAI model documentation. The Muse coordinate choice is based on recorded local search, not a claim of an official prescribed format.