© sisap challenge committee.
© sisap challenge committee.
The reproducibility of search methods is essential. We encourage sharing solutions using public GitHub repositories, preferentially under an open-source license to boost its usage and simplify its reproducibility by the community. Of course, repository documentation, notebooks, and tutorials are always welcome.
In the same terms, we ask for using GHA to ensure reproducibility under a limited-size dataset. GHA is a continuous integration platform that can run specified scripts after a repository is updated. Participants must ensure that indexing and searching methods work in one of the supported platforms. You can find small subset of the dataset allow checking your methods work. Below we provide examples of projects working with GHA for Python and Julia programming languages. Take a look on .github/workflows/ci.yml files, please feel free to use them as starting point for your solution.
https://github.com/sisap-challenges/sisap23-laion-challenge-faiss-example
SimilaritySearch.jlhttps://github.com/sisap-challenges/sisap23-laion-challenge-julia-example