We’ve made a lot of improvements to Model2Vec since it came out, many of which target the baseline performance of our distillation process. In this post, we walk through each change and explain why it matters for making your models smaller and faster.
Our newest shiny release is here! 0.3.8! This is a small release in line for a big one we’ll be releasing next week. See here for the release notes, and read on for details about ModernBERT compatibility (spoiler: it’s trickier than you’d think).
We’re super excited to announce the release of semhash, our semantic deduplication and dataset multitool (other features coming soon).
This blogpost describes the Tokenlearn method, which is a method to pre‐train Model2Vec models.
This blog was first posted on the Hugging Face blog. We’re also posting it here for archival purposes.
