NVIDIA/Megatron-Bridge/adding-model-support A
Guide for adding support for new LLM or VLM models in Megatron-Bridge.
Validate and use CUDA graph capture in Megatron Bridge, including local full-iteration graphs and Transformer Engine scoped graphs for attention, MLP, and MoE modules.
Install this skill (Claude Code)
# Clone and copy the skill into your project
git clone https://github.com/NVIDIA/skills.git
mkdir -p .claude/skills
cp -r skills/skills/Megatron-Bridge/perf-cuda-graphs .claude/skills/
# Or for personal use: ~/.claude/skills/Transparent heuristic — same formula for every entry. Total 85/100.
Guide for adding support for new LLM or VLM models in Megatron-Bridge.
Dev environment setup for Megatron Bridge — container-based development, uv package management, lockfile regeneration, adding dependencies, Slurm container usage, and common build...
Bump a pinned dependency (TransformerEngine, Megatron-LM, NRX, etc.), regenerate the lockfile, open a PR, and drive it to green by attaching a watchdog to the "CICD NeMo" workflow...
CI/CD reference for Megatron Bridge — pipeline structure, commit and PR workflow, CI failure investigation, and common failure patterns.
Code style and quality rules for Megatron Bridge — ruff configuration, naming conventions, type hints, mypy rules, docstrings, copyright headers, logging, and the code review check...
Run Megatron-LM (MLM) and Megatron Bridge training with mock or real data.