GSAI-ML/LLaDA-8B-Instruct

LLaDA-8B-Instruct

We introduce LLaDA, a diffusion model with an unprecedented 8B scale, trained entirely from scratch, rivaling LLaMA3 8B in performance.

Project Page

Code

Updates

[2025-10-21] We have modified modelingllada.py to support the input of attentionmask.