Fine-tune LLMs using the Tinker API. Covers supervised fine-tuning, reinforcement learning, LoRA training, vision-language models, and both high-level Cookbook patterns and low-level API usage.
安装
npx skills add sundial-org/skills --skill tinkernpx skills add sundial-org/skills --skill tinker -a claude-code -gnpx skills add sundial-org/skills --skill tinker -a codex -ggemini skills install https://github.com/sundial-org/skills.git --path skills/tinkergh skill install sundial-org/skills tinkergit clone --depth 1 https://github.com/sundial-org/skills /tmp/sundial-org__skills
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/sundial-org__skills/skills/tinker ~/.claude/skills/tinker # Claude Code
cp -R /tmp/sundial-org__skills/skills/tinker ~/.agents/skills/tinker # Codex, Cursor, Gemini CLI, Copilot, OpenCode