Orthrus - a chiennv Collection

chiennv 's Collections

Orthrus

updated about 10 hours ago

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

chiennv/Orthrus-Qwen3-4B

Text Generation • 5B • Updated about 12 hours ago • 22
chiennv/Orthrus-Qwen3-8B

Text Generation • 10B • Updated about 12 hours ago • 21
chiennv/Orthrus-Qwen3-1.7B

Text Generation • 2B • Updated about 12 hours ago • 37
Orthrus: Memory-Efficient Parallel Token Generation via Dual-View Diffusion

Paper • 2605.12825 • Published 3 days ago • 6