Orthrus - a chiennv Collection

chiennv 's Collections

Orthrus

updated 10 days ago

Fast, lossless LLM inference via dual-view diffusion decoding.

chiennv/Orthrus-Qwen3-4B

Text Generation • 5B • Updated 7 days ago • 583 • 8
chiennv/Orthrus-Qwen3-8B

Text Generation • 10B • Updated 7 days ago • 2.21k • 20
chiennv/Orthrus-Qwen3-1.7B

Text Generation • 2B • Updated 7 days ago • 1.08k • 8
Orthrus: Memory-Efficient Parallel Token Generation via Dual-View Diffusion

Paper • 2605.12825 • Published 12 days ago • 12