ai-post-training

Installation
SKILL.md

AI Post-Training

Domain: adaptation after pretraining, often after supervised fine-tuning — turning a base or SFT model into an aligned, preference-tuned, or reasoning-capable model with a reward signal. This skill owns the post-training decision and pipeline: when to post-train at all, which reward signal you can produce, which algorithm family fits, and how to keep it from over-optimizing. The per-algorithm catalogue and decision map (PPO, DPO, SimPO, KTO, ORPO, GRPO, GSPO, DAPO, RLOO, RLVR, RULER) live in references/methods-and-pipeline.md.

It does not cover: pretraining (ai-pretraining), the prompt→RAG→SFT promotion ladder (ai-architecture-advisor), or serving the result (ai-llm-inference).

Quick Reference

Installs
19
GitHub Stars
102
First Seen
Aug 12, 2026