cotalks.dev

GPU Memory Offload for LLM fine-tuning and inference with Phison aiDAPTIV+

(link)
note