Tencent Open-Sources AngelSpec: A Unified Training Framework for MTP and Block-Parallel Speculative Decoding on Hy3 Models

Tencent has open-sourced AngelSpec, a torch-native framework for training speculative-decoding draft models across six architectures, introducing DFly—a block-diffusion drafter with hybrid target conditioning and a hidden-correction autoregressive head—and integrating D-cut for runtime-adaptive verification budgeting. On the HY3-295B-A21B model with TP=8, DFly-8 achieves a 1.98–2.40× speedup over autoregressive decoding across concurrency levels 4 to 64.

Tencent has released AngelSpec, an open-source torch-native framework for training speculative-decoding draft models across six architectures. It introduces DFly, a block-diffusion drafter with hybrid target conditioning and a hidden-correction autoregressive head, and integrates D-cut for runtime-adaptive verification budgeting. On HY3-295B-A21B with TP=8, DFly-8 delivers a 1.98–2.40× speedup over autoregressive decoding across concurrency 4 to 64. The post Tencent Open-Sources AngelSpec: A Unified Training Framework for MTP and Block-Parallel Speculative Decoding on Hy3 Models https://www.marktechpost.com/2026/07/30/tencent-open-sources-angelspec-a-unified-training-framework-for-mtp-and-block-parallel-speculative-decoding-on-hy3-models/ appeared first on MarkTechPost https://www.marktechpost.com .