Unsloth: 2-5x faster LoRA/QLoRA fine-tuning, less VRAM.
安装
npx skills add NousResearch/hermes-agent --skill unslothnpx skills add NousResearch/hermes-agent --skill unsloth -a claude-code -gnpx skills add NousResearch/hermes-agent --skill unsloth -a codex -ggemini skills install https://github.com/NousResearch/hermes-agent.git --path optional-skills/mlops/training/unslothgh skill install NousResearch/hermes-agent unslothgit clone --depth 1 https://github.com/NousResearch/hermes-agent /tmp/NousResearch__hermes-agent
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/NousResearch__hermes-agent/optional-skills/mlops/training/unsloth ~/.claude/skills/unsloth # Claude Code
cp -R /tmp/NousResearch__hermes-agent/optional-skills/mlops/training/unsloth ~/.agents/skills/unsloth # Codex, Cursor, Gemini CLI, Copilot, OpenCode