# NandhaKishorM/vegaml

> Library for typed decisions from a frozen language model and a small physics engine: returns a choice, score, or probability with calibration, conformal sets, and an abstain flag. Supports 73k context and images, runs locally.

- Magnitude: 5.4 out of 10 — Early signal
- Stars: 121 total · +4 stars measured 2026-10-11, 02:45–04:06 UTC, ≈ 89 by evening
- Star trust: star growth looks organic
- Category: Language models · Language: Python · License: Apache-2.0 · Created: 2026-10-09 · Last push: 2026-10-11
- GitHub: https://github.com/NandhaKishorM/vegaml · Page: https://gitnova.dev/en/r/NandhaKishorM/vegaml

## Useful for

- Classify support emails by team with calibrated probabilities
- Score customer churn risk from message text without generating replies
- Label document types from images via decide_image

## Why it’s here

- Star-counter measurements on 2026-10-11 (UTC), 02:45–04:06: 117 → 121 stars (+4). This is the change over that interval.
- Estimated end-of-day forecast: about +89 stars, using observed gains and the previous day.
- The repository is 2 days old and already has 121 stars. With less than two weeks of history, there's no usual pace to compare the spike against yet.
- Top new repositories this week: #192.
- About 20 forks a day — people are taking the code.

## 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: 11
- Issues and pull requests: 12
- Watchers: 0
- Average over the last week: 70 per day
- Usual pace: too little history (under two weeks)
- Stars in the last hour (measured): 3
- Latest release: v0.8.0 (2026-10-11)

## Stars per day, last 3 days (oldest → newest, today is partial)

2026-10-09 … 2026-10-11: 0, 121, 4

## Spotted in now

- Top new repositories this week: #192

## Similar by description

1. **deepopen-com/deepopen** — 0.4 · Cooling · Language models · Python · +0 stars measured 2026-10-11, 00:06–04:16 UTC, ≈ 1 by evening
   Open-source non-autoregressive System 1 decision engine for typed decisions: returns typed answers (choice, score, noul) in a single forward pass across 100+ languages without generating text. Suited for classification, routing and…
   Full card: https://gitnova.dev/en/r/deepopen-com/deepopen.md
2. **NandhaKishorM/laya** — 3.0 · Cooling · AI agents · Python · +28 stars measured 2026-10-11, 00:40–04:09 UTC, ≈ 155 by evening
   Laya is a non-autoregressive decision engine that answers typed questions (choice, score, noul) over any state (text, email, ticket, JSON) in a single forward pass (~35 ms on GPU) with calibrated probabilities.
   Full card: https://gitnova.dev/en/r/NandhaKishorM/laya.md
3. **rkinas/basal** — 0.8 · Steady · Language models · Python · +0 stars measured 2026-10-11, 00:06–04:15 UTC, ≈ 3 by evening
   Inference engine for basal-1.0 — small, fast, calibrated typed-decision models for Polish and English. The model reads a state and returns a probability for each allowed answer in one forward pass, without generating text.
   Full card: https://gitnova.dev/en/r/rkinas/basal.md
4. **jiwidi/jiwo** — 1.6 · Steady · Language models · Python · +0 stars measured 2026-10-11, 00:02–04:12 UTC, ≈ 9 by evening
   Small decision models built on Qwen3.5 that return probabilities for answer options (choice, score, noul) in a single forward pass without generating text, with a median latency of about 50 ms on an H100.
   Full card: https://gitnova.dev/en/r/jiwidi/jiwo.md
5. **Bornholm/indecis** — 0.0 · Steady · Language models · Go · Not enough measurements today
   A Go library for training and running small decision models: a text encoder fine-tuned on your examples answers typed questions with calibrated probabilities instead of generated text.
   Full card: https://gitnova.dev/en/r/Bornholm/indecis.md

---

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 2026-10-11 04:17 UTC, updated every 30 minutes.