Show HN: Minimal LLM Post-Training Experiments on an 8GB GPU (SFT, DPO, GRPO)(github.com)7 points by popopanda 5 hours ago | 0 comments