prithivMLmods/Galactic-Qwen-14B-Exp2

Text Generation • Updated 6 days ago • 217 • 6

Note A 14B LLM currently ranks No. 3 on the leaderboard for 14B models, with an average score of 43.56. It achieved a score of 60 in Big Bench Hard (BBH 3-shot), securing the No. 2 spot among 14B models. Additionally, it holds the No. 2 position in MMLU-Pro with a score of 52.12 among 14B models. Uses SFT & Modular Combined Dataset