Tencent Hunyuan open-sources AngelSpec speculative decoding framework
Original title:🚀 We’ve open-sourced AngelSpec, an end-to-end speculative decoding framework supporting both traini...
AI summary
Tencent Hunyuan has open-sourced AngelSpec, an end-to-end speculative decoding framework supporting both training and deployment. On the Hy3-A21B model, its DFly scheme achieves 1.98–2.40x end-to-end speedup over autoregressive decoding, with throughput 10.5–11.8% higher than DFlash. The training code and Hy3-A21B MTP/DFly draft model weights have been open-sourced.
Key Takeaways
- Tencent Hunyuan has open-sourced AngelSpec,an end-to-end speculative decoding framework supporting…
- On the Hy3-A21B model,its DFly scheme achieves 1.
- 40x end-to-end speedup over autoregressive decoding,with throughput 10.
- 8% higher than DFlash.
- The training code and Hy3-A21B MTP/DFly draft model weights have been open-sourced.
Quick Facts
- Published
- Updated
- Source
- X:腾讯混元 (@TencentHunyuan)
- Category
- Products & Apps
- AI score
- 74
做推理加速的可以试试这个端到端推测解码框架,腾讯开源了完整代码和权重,在 Hy3 上提速近 2.4 倍且吞吐更高,拿来即用。
Related entities
Source & copyright
This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.
How to Cite
When citing this article, keep the title, date, and permalink, and credit Orange News as the source.
Orange News: Tencent Hunyuan open-sources AngelSpec speculative decoding framework. July 29, 2026. https://new.nat6.net/en/news/cms639jt416x6robksy69dzel/