CommunityJev Models & Reimplementations
Train your own decision model with Unsloth
Unsloth's guide to turning a generative LLM into a Jev-style decision model that returns a choice with a probability instead of text, plus a companion page for serving Laya behind the TypeSafe API.
- Published by
- Community
- Last verified
- Oct 2026
- Languages
- Python
- Capabilities
- Open sourceRuns locally
Highlights
- Fine-tunes Qwen and Gemma LLMs into decision models that score options and return a choice with a probability.
- Reports Qwen3.5 0.8B moving from 20.7% to 74.3% across three decision benchmarks on about 4GB of VRAM.
- Uses a Clef-style head with LoRA (r=64) for one epoch, with downstream accuracy reported from 30-37% to 78%.
- A companion doc serves Laya and Clef behind a TypeSafe-compatible local API so existing Jev clients work unchanged.
Watch out
The figures are Unsloth's own and measure open models imitating the decision shape, not TypeSafe's hosted Jev; treat the accuracy claims as a starting point to reproduce.
Questions about this entry
Short answers drawn from the linked project or page.
A decision model from an ordinary generative LLM: you fine-tune a model like Qwen or Gemma so it scores the options you provide and returns a choice with a probability instead of writing text.
It reports Qwen3.5 0.8B rising from 20.7% to 74.3% aggregate accuracy across three decision benchmarks, trained on about 4GB of VRAM, and a Clef-head LoRA run going from 30-37% to 78%.
The release notes describe a Clef-style decision head with LoRA at r=64 for a single epoch on top of the base model, so the base weights stay mostly frozen and the run fits on modest hardware.
Yes. A companion Unsloth page exposes Laya and Clef through a TypeSafe-compatible local API, so existing Jev clients work unchanged once you point TYPESAFE_BASE_URL at the local server.
No. Jev is a hosted, closed model. This trains open models to reproduce the decision shape, a choice with probabilities, and the reported numbers are Unsloth's own, not TypeSafe's.
The guide is built around Unsloth's local tooling, including a desktop app for macOS, Windows, and Linux and an install script, so the data can stay on your machine.
The documentation names Qwen3.5 and Gemma 4 among the convertible bases, and the release notes frame it as turning any text or vision LLM into a Jev-style decision model.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
LinkedIn: What multimodal decision models actually change, and what they do not.
LinkedInJevDirectory.org
Hacker News: Decision Index 0.2.1 scores open reproductions of Jev on a frozen suite.
jackbravo4 pointsDecision Index 0.2.1 – 70 open reproductions of TypeSafe Jev's Decision ModelRead the thread on Hacker NewsHacker News: Pointer to a DiffusionGemma plus vLLM tutorial for a Jev-shaped classifier.
1 pointHow to build a Jev-style classifier with DiffusionGemma and vLLMRead the thread on Hacker NewsHacker News: A one-token logprob wrapper that also classifies webcam frames.
allanrbo151 pointsA single function Jev-like wrapper for LLMs, including vision modelsRead the thread on Hacker NewsMore Jev resources like this
59GitHub stars
A preregistered measurement of a Jev-style interface on frozen Qwen3-4B: read the option letter's logits instead of generating grammar-constrained JSON.
Jev Models & Reimplementations#community#python#local
Community
r-ms/mini-jevPython
r-ms/mini-jevPython6.9kGitHub stars
A native MLX runtime for Laya's open typed-decision weights: short decisions on Apple Silicon, with no generated tokens, PyTorch, or cloud API.
Jev Models & Reimplementations#community#python#open-models
Community
mizorewww/laya-mlxPython
mizorewww/laya-mlxPython4.8kGitHub stars
An open-model semantic-if engine positioned as a local Jev-style decision stack, separate from the original SemIf repo but sharing the same author.
Jev Models & Reimplementations#community#python#open-models
Community
TheoLeeCJ/SemIf-OpenJevPython
TheoLeeCJ/SemIf-OpenJevPython1.6kGitHub stars
Laya typed decisions through Apple Core ML and the Neural Engine, with the author's speed and energy benchmarks for short decisions on an M3 Max.
Jev Models & Reimplementations#community#python#open-models
Communitymizorewww/laya-coremlPython
mizorewww/laya-coremlPython1.3kGitHub stars
A local yes/no decision model that speaks Jev's wire format on a laptop CPU, refusing Choice and Score until those question types ship.
Jev Models & Reimplementations#community#python#open-models
Community
feder-cr/jevC++
feder-cr/jevC++877GitHub stars
A local server that reads next-token probabilities from a fine-tuned Spark model and returns Choice, Score, and Noul answers with zero generated tokens.
Jev Models & Reimplementations#community#python#open-models
Community
Rizzo-AI-Academy/rizzo-flowPython
Rizzo-AI-Academy/rizzo-flowPythonFrom the Jev guides
Original write-ups on what this entry is for.
Base decision checkpoints sit near chance on purpose. The typed-decisions loop, the corpora behind the students, and the temperature fit that turns probabilities into thresholds.
Jev Practices & Patterns#fine-tuning#calibration#open-models
Read article
Jev takes text only; Clef-omni adds image, audio, and video input and Liquid d1 adds images, each keeping the same typed-answer shape. Where that matters, where it does not, and the costs the vendors underplay.
Jev Models & Reimplementations#multimodal#open-models#decision-model
Read article
One model in September, a dozen by October. Clef, Liquid d1, Perplexity, and OpenAI all ship the same typed-answer shape now. How to read the rival benchmarks and pick on your own task.
Jev Models & Reimplementations#decision-model#benchmarks#pricing
Read article
From the Jev community
Posts from builders shipping with Jev right now.