{"name":"Soup","repo":"MakazhanAlpamys/Soup","url":"https://midorreal.com/project/makazhanalpamys-soup","repository":"https://github.com/MakazhanAlpamys/Soup","description":"Fine-tune LLMs from one YAML. Layer streaming trains an 8B model on a 4 GB laptop GPU.","facts":{"license":{"value":"Apache-2.0","source":"https://github.com/MakazhanAlpamys/Soup","checked":"22 Nov 2026"},"language":{"value":"Python","source":"https://github.com/MakazhanAlpamys/Soup","checked":"22 Nov 2026"},"homepage":{"value":"https://trysoup.dev","source":"https://trysoup.dev","checked":"22 Nov 2026"},"last_release":{"value":"v0.78.0, 18 Nov 2026","source":"https://github.com/MakazhanAlpamys/Soup/releases","checked":"22 Nov 2026"},"releases":{"value":"185","source":"https://github.com/MakazhanAlpamys/Soup/releases","checked":"21 Nov 2026"},"contributors":{"value":"165","source":"https://github.com/MakazhanAlpamys/Soup/graphs/contributors","checked":"21 Nov 2026"},"archived":{"value":"no","source":"https://github.com/MakazhanAlpamys/Soup","checked":"22 Nov 2026"}},"stars":10718.0,"writeup":{"what_it_is":"Soup is a command-line tool for fine-tuning and post-training LLMs from one YAML config file. It includes layer streaming, which the README says lets an 8B model train on a 4 GB laptop GPU, and it also provides chat, merge, export and push commands.","audience":"Developers and ML practitioners who fine-tune LLMs locally or on their own GPUs","claims":[{"kind":"broad","claim":"Fine-tunes LLMs from a single YAML config and command","excerpt":"One config, one command, done.","status":"too_broad"},{"kind":"specific","claim":"Layer streaming trains an 8B model on a 4 GB laptop GPU","excerpt":"Fine-tune an 8B model on a 4 GB laptop GPU.","status":"not_checked"},{"kind":"specific","claim":"Measured 119.6 tok/s at 3.32 GB peak on an RTX 3050 Laptop 4 GB (pre-repair)","excerpt":"119.6 tok/s at 952 MHz, 3.32 GB peak","status":"not_checked"},{"kind":"specific","claim":"Exports to GGUF for Ollama and llama.cpp","excerpt":"GGUF for Ollama / llama.cpp","status":"not_checked"}],"alternatives":[],"written_by":"AI, from the project's README","date":"2026-11-21"},"why_now":null,"owner_supplied":null,"score":{"verdict":"real","hype":80,"reality":88,"gap":-8,"momentum":0.5,"confidence":85,"version":"score-2.0","rule":"verdict-2.0","calculated_at":"2026-11-22T08:00:00+00:00","verdict_source":"rule"},"early_read":null,"changes":[{"date":"2026-11-18T18:29:45+00:00","text":"Latest release v0.78.0","source":"https://github.com/MakazhanAlpamys/Soup/releases"},{"date":"2026-03-03T19:03:29+00:00","text":"First release","source":"https://github.com/MakazhanAlpamys/Soup/releases"},{"date":"2026-02-20T11:05:01+00:00","text":"Repository created","source":"https://github.com/MakazhanAlpamys/Soup"}],"cite":"Soup is a GitHub project at https://github.com/MakazhanAlpamys/Soup written mainly in Python, described there as \"Fine-tune LLMs from one YAML. Layer streaming trains an 8B model on a 4 GB laptop GPU\".","attribution":"Data from Mid or Real, https://midorreal.com/project/makazhanalpamys-soup, checked 22 November 2026. Attribution with a link is required.","still_real":null}