Fine-tune and post-train language models (SFT, DPO, LoRA and QLoRA) from a single YAML file and one command, on consumer GPUs.
Made by MakazhanAlpamys, Apache-2.0 licensed.
Honorary mentionLiveOpen source
Soup
Fine-tune LLMs from one YAML file; layer streaming trains an 8B model on a 4 GB laptop GPU.