MinishLab/model2vec
View on GitHub ↗ ★ 2,264 Python
Fast State-of-the-Art Static Embeddings
8 mentions on Hacker News by 5 people: 5 stories, 3 comments · first 2024-09-22 · latest 2026-10-14
20082026
-
…I think it loses the order of words but, someone's gotta try and see if it works anyway right? https://github.com/MinishLab/model2vec
comment by Ey7NFZ3P0nzAe · 2026-10-14 · in We turned EmbeddingGemma 2 into a speech-to-text model -
Can anyone test it through model2vec? https://github.com/MinishLab/model2vec
comment by Ey7NFZ3P0nzAe · 2025-09-04 · in EmbeddingGemma: The Best-in-Class Open Model for On-Device Embedding -
…a bit bigger (~100mb). There's definitely a quality trade-off. We have extensive benchmarks here: https://github.com/MinishLab/model2vec/blob/main/results/REA... . potion-base-32M reaches ~92% of the performance of MiniLM while being much…
comment by Tananon · 2025-05-18 · in Show HN: Model2vec-Rs – Fast Static Text Embeddings in Rust