sparse-lavida-diffusion

【效率与协作】Accelerate masked discrete diffusion models by dynamically truncating unnecessary masked tokens while maintaining complete representation through positional information. Use register tokens as compressed representations, implement step-causal attention masks for bidirectional context. Achieve 1.95-2.83× speedups across generation tasks.

独立 GitHub 快照 · 自动分类 · 以原作者文档为准