The mission of sotabench is to provide an insight into the performance of open source Machine Learning models. Sotabench is a twin website to Papers with Code, with Papers with Code focusing on results reported in papers, and sotabench focusing on results obtained by continously running code libraries.
We provide a free benchmarking service for all open source ML repositories that is integrated with GitHub. It is like Continuous Integration, but instead of running unit tests, it runs benchmarks.
The service is open, anyone can add a benchmark and anyone can add their model. Anyone can show their implementation reproduces a paper and can compare to other repositories on equal footing.