tomaarsen
@tomaarsen
Sentence Transformers Lead, SetFit & NLTK maintainer
Machine Learning Engineer at ๐ค Hugging Face
511 Following 5K Followers
๐งต I've just released Sentence Transformers v6.2.0!
Featuring model merging with @arcee_ai mergekit, load models faster, & evaluators share work across GPUs automatically.
Plus lower-memory losses, multi-vector improvements & plenty of fixes.
Thread with highlights ๐
Show more
๐@Google's EmbeddingGemma2 is looking to be a huge success!
After only a week, it has surpasses EG1's likes to become the 4th most liked Google model ever, as well as the 3rd most liked Sentence Transformers model ever.
Congratulations to the team for their work!
Links in ๐งต
Show more
Very cool: Tencent has updated the license of their SOTA multimodal embedding models to Apache-2.0!
Tencent released WeMM-Embedding this week with a technical report; it's the top trending paper on Papers with Code
The models outperform Gemini Embedding 2, Qwen3-VL Embedding, and @VoyageAI embedding models
Show more
BIG ANNOUNCEMENT FROM HUGGING FACE TODAY:
We're unveiling Microduck ๐ฅ๐ค
It's a tiny $399 open-source robot you can teach new tricks with reinforcement learning. It can walk, pick things up, get back up when it falls, and even roller-skate.
Welcome to the era of open-source affordable robots to democratize physical AI and world models!
๐ค๐ค๐ค
Show more
Summer of open-source ain't over yet ๐ฅ๐ฅ๐ฅ
๐ New blog post: Training and Finetuning Multi-Vector Embedding Models with Sentence Transformers.
As a practical example, I finetuned a ColBERT-style model for medical retrieval. 14.5 hours on one RTX 3090, and it beats every general-purpose retriever I could find.
Thread ๐งต
Show more
At 1400x cheaper, I know I'm sticking to embeddings (dense, sparse, multi-vector), plus hybrid (incl. bm25) and rerankers.
Perhaps I'd even use listwise cross-encoders, they seem interesting.
Regarding Sentence Transformers v6.0 update:
A dense model compresses a whole text into one vector, then compares two vectors.
A multi-vector model keeps one vector per token, scores every query token against every document token, takes the best match for each, and sums those.
Show more
@jeremyphoward @answerdotai He's right! 200k monthly downloads and not for nothing, here's the link for those who want to try it:
๐จI've just released Sentence Transformers v6.0!
MultiVectorEncoder joins the family: ColBERT-style late interaction models are now a first-class model type, for training, inference & interpretation, alongside dense, sparse & reranker models.
Big thread ๐งต
Show more
