Soup CLI: Fine-Tune an 8B LLM on a 4GB Laptop GPU, No Cloud Required (2026 Guide)
Soup CLI is a free, open source LoRA fine-tuning tool that streams frozen base model layers from RAM to GPU one at a time. Train Llama-3.1-8B at 119.6 tok/s in 3.32GB peak VRAM. SFT, DPO, GRPO, KTO.
8 min readRead more →