Fixed legal_moves bug and skipped already evaluated models faa67d0 nathanael-fijalkow commited on 9 days ago
First version ready with webhook and deterministic eval cb44915 nathanael-fijalkow commited on 11 days ago
Integrated the template, made evaluation more robust to different tokenizers 8a7719b nathanael-fijalkow commited on 17 days ago
removed interactive demo and improved username finding aee63fa nathanael-fijalkow commited on 19 days ago