megatron-lm-20210305


  • [Megatron](https://arxiv.org/pdf/1909.08053.pdf) is a large, powerful transformer developed by the Applied Deep Learning Research team at NVIDIA. This repository is for ongoing research on training large transformer language models at scale. We developed efficient, model-parallel (tensor and pipeline), and multi-node pre-training of [GPT](https://arxiv.org/abs/2005.14165) and [BERT](https://arxiv.org/pdf/1810.04805.pdf) using mixed precision.