Smarter, Better, Faster, Longer: A Modern Bidirectional Encoder for Fast, Memory Efficient, and Long Context Finetuning and Inference

Introduces ModernBERT, a bidirectional encoder advancing BERT-like models with 8K context length.

📝 https://arxiv.org/abs/2412.13663
👨🏽‍💻 https://github.com/AnswerDotAI/ModernBERT

Comments