大模型微调速度提升 2 倍、显存减少 70%，单卡也能训练。
Transformer Engine
A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit and 4-bit floating point (FP8 and FP4) precision on Hopper, Ada and Blackwell GPUs, to provide better performance with lower memory util
- 仓库
- NVIDIA/TransformerEngine
- Stars
- ★ 3,576
- Forks
- 861
- 语言
- Python
- 许可证
- Apache-2.0
- 最近更新
- 2026-10-12
- 访问次数
- 0
PyTorch
