# MakazhanAlpamys/Soup

> A CLI tool for fine-tuning and post-training LLMs from a single YAML config, with layer streaming that lets an 8B model train on a 4 GB laptop GPU.

- Magnitude: 2.4 out of 10 — Cooling
- Stars: 9,636 total · +13 stars measured 2026-10-25, 00:13–11:05 UTC, ≈ 28 by evening
- Star trust: star growth looks organic
- Category: Language models · Language: Python · License: Apache-2.0 · Created: 2026-02-20 · Last push: 2026-10-24
- GitHub: https://github.com/MakazhanAlpamys/Soup · Homepage: https://trysoup.dev · Page: https://gitnova.dev/en/r/MakazhanAlpamys/Soup

## Useful for

- Run QLoRA fine-tuning from a single soup.yaml without setting up infrastructure
- Train an 8B model locally on a 4 GB laptop GPU via stream_layers
- Validate a config with soup doctor before starting a training run

## Why it’s here

- Star-counter measurements on 2026-10-25 (UTC), 00:13–11:05: 9623 → 9636 stars (+13). This is the change over that interval.
- Estimated end-of-day forecast: about +28 stars, using observed gains and the previous day.
- Interest is fading: the two-day pace is 30% of the previous week and a half.
- About 12 forks a day — people are taking the code.
- Recent forks include notable developers: @SauravMukherjee44 (975 followers).

## Star trust

Star growth looks organic.

Star-trust labels are heuristics based on the repository’s behavior, not a check of every stargazer.

## Numbers

- Forks: 1,575
- Issues and pull requests: 2,168
- Watchers: 59
- Average over the last week: 43 per day
- Usual pace: 95 per day
- Latest release: v0.76.1 (2026-10-14)

## Stars per day, last 30 days (oldest → newest, today is partial)

2026-10-04 … 2026-11-02: 68, 75, 79, 47, 49, 65, 62, 65, 65, 286, 134, 73, 79, 81, 83, 70, 54, 49, 35, 37, 27, 5, None, None, None, None, None, None, None, None

## Magnitude by day, last 14 days

10-12 2.7, 10-13 2.8, 10-14 4.4, 10-15 3.7, 10-16 3.0, 10-17 2.9, 10-18 3.1, 10-19 3.2, 10-20 2.8, 10-21 2.5, 10-22 2.4, 10-23 2.1, 10-24 2.5, 10-25 2.4

---

Magnitude (0–10) measures how fast and how unusually interest in a repository is growing right now. It is not a quality score.
Days are UTC. “So far today” is a fact; “expected by the end of the day” is a forecast.
Summaries and use cases are written by an LLM (DeepSeek V4.1 Flash) from the README and may be inaccurate: verify specific claims (benchmarks, speed, hardware) in the repository itself.
Data as of 2026-10-25 11:13 UTC, updated every 30 minutes.