🔥Microsoft Open-Source Embedding Model Harrier Tops Multilingual MTEB Leaderboard, Leading OpenAI and Google


According to 1M AI News monitoring, Microsoft Bing team has open-sourced the embedding model series Harrier. Embedding models are the foundational components of search engines and RAG systems, responsible for converting text into vectors for retrieval and matching. The flagship Harrier-OSS-v1-27B scored an average of 74.3 on the multilingual MTEB v2 benchmark (covering 131 tasks), surpassing the previous highest score of open-source models by 2 percentage points, ranking first. Compared to closed-source models, OpenAI's text-embedding-3-large scored an average of 58.92, and Google's Gemini Embedding 2 scored 69.9, Amazon Titan Embed v…
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments
  • Pin