Offline
day
week
Show HN: Minimal LLM Post-Training Experiments on an 8GB GPU (SFT, DPO, GRPO)
15 points
|
github.com
|
popopanda
|
8hrs