Ornith-1.5 Open-Sources DFlash Draft Models for Up to 2.54x Lossless Speedup

ornith-ai has released DFlash draft model checkpoints for the full Ornith-1.5 family (9B, 35B-A3B, 397B) on Hugging Face, enabling speculative decoding with up to 2.54x lossless inference speedup.

2026-09-29 ~ 2026-10-01 · 2 related posts

• Ornith-1.5 releases DFlash draft checkpoints for 9B/397B/35B-A3B speculative decoding — jacek2023 · 2026-09-29
• Ornith-1.5 DFlash draft models deliver up to 2.54x lossless inference speedup — alan_ritter · 2026-10-01