💡 What's new in txtai 9.0
The 9.0 release adds first class support for sparse vector models (i.e. SPLADE), late interaction models (i.e. ColBERT), fixed dimensional encoding (i.e. MUVERA) and reranking pipelines ✨ The embeddings framework was overhauled to seamlessly support ...
Responses
Markdown supported · ⌘↵ to post
Anik Sikder
Turning bugs into features since forever
This release really puts txtai in the same conversation as top-tier retrieval frameworks. The addition of SPLADE and ColBERT support is huge especially for anyone working on high-precision semantic search or RAG pipelines. MUVERA’s fixed dimensional encoding feels like the missing piece for scaling multi-vector models without sacrificing performance. Curious to see how this impacts latency and recall in real-world use cases.