AngelSpec
Machine Learning Model TrainingML Frameworks

AngelSpec

A unified, torch-native training framework for MTP and block-parallel speculative decoding.

Open Source

About

AngelSpec is developed by the Tencent Hunyuan AI Infra team, which is a torch-native framework for training speculative-decoding draft models, covering both autoregressive MTP drafting and the block-parallel DFlash family. It is the training framework behind the technical report: all drafters in the report — the TTT-trained MTP drafter and the DFly family — are trained and released with it.

Open Source Health

Not enough history
Stars
266
Forks
29
License
Not stated
Last commit
2 months ago
Python

Related Categories