Xiuyu Li(@sheriyuo):SLAI T-Rex raises full-parameter DeepSeek-V4-Pro post-training on Ascend from 11.67% to 34.22% MFU, then demonstrates a solver-grounded CPT and SFT pipeline on the Flash model. SLAI T-Rex: Full-Parameter Post-training of the DeepSeek-V4 Family on Ascend SuperPOD Paper:

Xiuyu Li

@sheriyuo

Researcher @StepFun_ai | Working on long-horizon tasks | Prev @RUC1937 | Opinions are my own

加入 February 2026

1.7K 正在關注 12.4K 粉絲

Xiuyu Li@sheriyuo

2026.07.23 04:22

SLAI T-Rex raises full-parameter DeepSeek-V4-Pro post-training on Ascend from 11.67% to 34.22% MFU, then demonstrates a solver-grounded CPT and SFT pipeline on the Flash model. SLAI T-Rex: Full-Parameter Post-training of the DeepSeek-V4 Family on Ascend SuperPOD Paper: