Charles Pierse

RT @drdannywilliams: Late chunking - a paper I co-authored with @JinaAI_ - shows that switching the order in which you chunk and embed you…

Really impressed by this approach and how well this new model performs on benchmarks esp given its relatively small parameter count. More importantly, I love this trend of investing in new approaches and research for embedding models. LLM's have gotten a lot of love for the