training-llms-megatron
training-llms-megatron is a Model training & evaluation skill for AI agents, published by davila7 in the claude-code-templates repository on GitHub. Skillspace rates training-llms-megatron B (84/100) and it passed the safety scan; a judge model has also reviewed it.
Trains large language models (2B-462B parameters) using NVIDIA Megatron-Core with advanced parallelism strategies. Use when training models >1B parameters, need maximum GPU efficiency (47% MFU on H100), or require tensor/pipeline/sequence/context/expert parallelism. Production-ready framework used for Nemotron, LLaMA, DeepSeek.
- Grade
- B
- Quality
- 84 / 100
- Safety
- Passed
- GitHub stars
- 30,694
Placed #1 of 6 in the judged category final for Model training & evaluation.
This rating describes the version of the repository we crawled; it may have changed since, so go by its current files.
Our review
- Architecture
- 94 / 100
- Maturity
- Packaged: licensed, with a version or references
- Security scan
- No issues found
- License
- Not declared. Check the repository before reusing it.
Judge review
Placed #1 of 6 in the judged category final for Model training & evaluation.
- Trigger
82 82 - Instructions
78 78 - Specificity
88 88 - Safety
80 80 - Disclosure
45 45 - Maintenance
42 42
Concrete parallelism tables and real H100 MFU numbers make this the most specific entry, though it's one long file with no reference-based disclosure.
Not checked yet
- Category is a classifier prediction; no model has read this skill yet
- No blind task eval yet
- Scope (portable / vendor / repo-internal) not assessed; only the top of each category is read for scope
- Sibling files (scripts, references) not fetched; security scan covers SKILL.md only
Skillspace indexes and rates this skill; it does not host or endorse it. Run third-party skills at your own risk.