M
Megatron-LM
NVIDIA/Megatron-LM
Ongoing research training transformer models at scale
★17.4kstars
Python
NOASSERTION
Updated: Today
📋 Project at a Glance
Tap to expand
What's this?A large-language-models/model-para tool in the Data & Infrastructure category, built with Python, open-source
Who made it?Maintained by NVIDIA team, 17.4K⭐ on GitHub, #106 out of 3133 in Data & Infrastructure
Why does it exist?As the Data & Infrastructure landscape evolved, the NVIDIA team identified the need for better large-language-models solutions. Megatron-LM was created to simplify model-para workflows.
What can it do?Key use cases: transformers
How to install with AI?Use an AI coding assistant (Claude Code, Cursor, Copilot) to automatically set up pip dependencies and virtual env. Follow the README — the AI handles the rest.
🔗 github.com/NVIDIA/Megatron-LM | 官网 https://docs.nvidia.com/megatron-core/developer-guide/latest/get-started/quickstart.html
🔗 github.com/NVIDIA/Megatron-LM | 官网 https://docs.nvidia.com/megatron-core/developer-guide/latest/get-started/quickstart.html
Topics
large-language-modelsmodel-paratransformers