Glossary: Serving LLMs at Scale
EAGLE / MTP
Also written: EAGLE, MTP, multi-token prediction
Extra small layers trained to predict several tokens ahead, used as a drafter. Some models, including Qwen3.5, ship with them.
Glossary: Serving LLMs at Scale
Also written: EAGLE, MTP, multi-token prediction
Extra small layers trained to predict several tokens ahead, used as a drafter. Some models, including Qwen3.5, ship with them.