distributed-training-deepspeed
Agent skill fromdavila7/claude-code-templates
Expert guidance for distributed training with DeepSpeed - ZeRO optimization stages, pipeline parallelism, FP16/BF16/FP8, 1-bit Adam, sparse attention
How to install distributed-training-deepspeed
install distributed-training-deepspeed
# install the claude-code-templates pack with the skills CLI
npx skills add davila7/claude-code-templates# or add only this skill to Claude Code: copy the folder
cli-tool/components/skills/ai-research/distributed-training-deepspeed/
# into
~/.claude/skills/distributed-training-deepspeed/ (all projects)
.claude/skills/distributed-training-deepspeed/ (one project)
1# Deepspeed Skill23Comprehensive assistance with deepspeed development, generated from official documentation.45## When to Use This Skill67This skill should be triggered when:8- Working with deepspeed9- Asking about deepspeed features or APIs10- Implementing deepspeed solutions11- Debugging deepspeed code12- Learning deepspeed best practices1314## Quick Reference1516### Common Patterns17…
Excerpt shown under the repository's MIT license.
