Join the conversation

Join the community of Machine Learners and AI enthusiasts.

Sign Up
m-ricΒ 
posted an update 1 day ago
Post
627
After 6 years, BERT, the workhorse of encoder models, finally gets a replacement: π—ͺ𝗲𝗹𝗰𝗼𝗺𝗲 π— π—Όπ—±π—²π—Ώπ—»π—•π—˜π—₯𝗧! πŸ€—

We talk a lot about ✨Generative AI✨, meaning "Decoder version of the Transformers architecture", but this is only one of the ways to build LLMs: encoder models, that turn a sentence in a vector, are maybe even more widely used in industry than generative models.

The workhorse for this category has been BERT since its release in 2018 (that's prehistory for LLMs).

It's not a fancy 100B parameters supermodel (just a few hundred millions), but it's an excellent workhorse, kind of a Honda Civic for LLMs.

Many applications use BERT-family models - the top models in this category cumulate millions of downloads on the Hub.

➑️ Now a collaboration between Answer.AI and LightOn just introduced BERT's replacement: ModernBERT.

π—§π—Ÿ;𝗗π—₯:
πŸ›οΈ Architecture changes:
β‡’ First, standard modernizations:
- Rotary positional embeddings (RoPE)
- Replace GeLU with GeGLU,
- Use Flash Attention 2
✨ The team also introduced innovative techniques like alternating attention instead of full attention, and sequence packing to get rid of padding overhead.

πŸ₯‡ As a result, the model tops the game of encoder models:
It beats previous standard DeBERTaV3 for 1/5th the memory footprint, and runs 4x faster!

Read the blog post πŸ‘‰ https://huggingface.co/blog/modernbert
In this post