Soup CLI
Fine-tune an 8B LLM on a 4 GB laptop GPU
Soup CLI launched on August 9, 2026 with 107 votes, #336 of 23,274 launches that month and more than 95% of that year's launches.
What it does
LoRA keeps the base model frozen: read, never written. So Soup keeps it in system RAM and streams it into the GPU one decoder layer at a time. Peak VRAM becomes one layer instead of the whole model. Measured on an RTX 3050 Laptop 4 GB: Llama-3.1-8B trains at 119.6 tok/s in 3.32 GB peak. One YAML, one command. SFT, DPO, GRPO, KTO, plus eval, gating and export. Apache-2.0. Every number is published, including the ones I measured and threw away.
Fine-tune Llama-3.1-8B on a 4 GB laptop GPU, and align on the same card. Soup is at v0.76.1, tagged 14 October 2026, a maintainer patch on the v0.76.0 line: the bundled DPO examples train again, every YAML file Soup reads goes through one loader with nesting and expansion limits and a size cap per reader (a soup.yaml over 1 MiB or nested deeper than 100 levels stops the load with one line naming t
Does a similar job
all alternatives →
80% faster, 50% less memory, 0% loss of accuracy Llama finetuning2023 · github.com · ▲385Hi HN! I'm just sharing a project I've been working on during the LLM Efficiency Challenge - you can now finetune Llama with QLoRA 5x faster than Huggingface's original implementation on your own local GPU. Some highlights: 1. Manual autograd engine - hand derived backprop steps. 2. QLoRA / LoRA 80% faster, 50% less memory. 3. All kernels written in OpenAI's Triton language. 4. 0% loss in accuracy - no approximation methods - all exact. 5. No change of hardware necessary. Supports NVIDIA GPUs since 2018+. CUDA 7.5+. 6. Flash Attention support via Xformers. 7. Supports 4bit and 16bit…
Llama 3.1 70B on a single RTX 3090 via NVMe-to-GPU bypassing the CPUFeb 2026 · github.com · ▲395Hi everyone, I'm kinda involved in some retrogaming and with some experiments I ran into the following question: "It would be possible to run transformer models bypassing the cpu/ram, connecting the gpu to the nvme?" This is the result of that question itself and some weekend vibecoding (it has the linked library repository in the readme as well), it seems to work, even on consumer gpus, it should work better on professional ones tho
Finetune LLaMA-7B on commodity GPUs using your own text2023 · github.com · ▲449I've been playing around with https://github.com/zphang/minimal-llama/ and https://github.com/tloen/alpaca-lora/blob/main/finetune.py, and wanted to create a simple UI where you can just paste text, tweak the parameters, and finetune the model quickly using a modern GPU. To prepare the data, simply separate your text with two blank lines. There's an inference tab, so you can test how the tuned model behaves. This is my first foray into the world of LLM finetuning, Python, Torch, Transformers, LoRA, PEFT, and Gradio. Enjoy!
Fine-tune an 8B model on a 4 GB laptop GPUAug 2026 · github.com · ▲139Fine-tune LLMs from one YAML. Layer streaming trains an 8B model on a 4 GB laptop GPU. - MakazhanAlpamys/Soup
Run Llama3.1 405B on a 8GB VRAM challenge [video]2024 · youtube.com · ▲19How to run Llama3.1 405B on a 8GB VRAM
KVSplit – Run 2-3x longer contexts on Apple Silicon2025 · github.com · ▲272I discovered that in LLM inference, keys and values in the KV cache have very different quantization sensitivities. Keys need higher precision than values to maintain quality. I patched llama.cpp to enable different bit-widths for keys vs. values on Apple Silicon. The results are surprising: - K8V4 (8-bit keys, 4-bit values): 59% memory reduction with only 0.86% perplexity loss - K4V8 (4-bit keys, 8-bit values): 59% memory reduction but 6.06% perplexity loss - The configurations use the same number of bits, but K8V4 is 7× better for quality This means you can run LLMs with 2-3× longer…
More ai this month
the category →
Paintings by AI models: programs run through a simulation of oil paint.
AI · 20d ago · stillwet.art
PixVerse R2▲381A real-time world model you can explore and change
AI · 28d ago · world.pixverse.video
Source: Product Hunt launch ↗
Launched alongside, August 2026
the whole month →- TL
Life & fun · Aug 2026 · louisabraham.github.io
Clipto MCP▲655Hey HN, We recently built Clipto. It’s a tool that lets you search over terabytes of video, audio, and images on your computer, without relying on the cloud. Motivation: we probably all had this similar experience, we know a moment exists in a video or audio, but finding it takes hours scrubbing the timeline. You can send all the media to process in the cloud, but it’s slow, expensive and raises privacy concerns. So we decided to build our own on-device media search engine. How it works (high level): 1. We ingest video, audio, image; normalize formats via ffmpeg; run content analysis to…
Work · Aug 2026 · clipto.com
Hello HN! I found that picking out plausible but diverse skin tones for my digital art and game development projects was kind of difficult, and I got curious about if there was a way to define a color space that made it easy. I've built a color picker and procedural generation algorithm based on the space as well as a bunch of other fun js features and demos throughout the page that use the equations. If you find it interesting, I have lots of explanations of how I built it and what properties the space has. The methodology might be a bit shaky, but hopefully the result is as helpful for…
Life & fun · Aug 2026 · toneyalexander.github.io
Astute▲606Automate your B2B brand going viral, with new media creators
AI · Aug 2026 · company-app.joinastute.com