jev-curate

High-throughput synthetic & pretraining dataset sifter powered by TypeSafe AI Jev (api.

What it does

High-throughput synthetic & pretraining dataset sifter powered by TypeSafe AI Jev (api.typesafe.ai). Stream, filter, and score Parquet & JSONL datasets at 1,500+ rows/sec using System One typed decisions (Choice, Score, Noul).

Numbers, harnesses, and comparisons you can rerun on your hardware instead of trusting a screenshot. Tags in the wild: benchmarks. Skim the repo for setup, then open the demo when you want proof before you merge anything.

Related products

Hand-picked neighbors with rich profiles or overlapping tags.

BenchmarksPython4,785

DemoFeatured

SemIf

Semantic ifs on a home 3090: open models doing structured branches without pretending to be TypeSafe.

benchmarksopen-sourceself-hosted

Details Demo

BenchmarksPython851

Featured

jevlike

Train a small model that chooses among a changing list of text options, one probability per option in a single pass.

by @hhkkmon

benchmarksgame

Details

BenchmarksTypeScript31

Demo

typesafe-ai-benchmark

A gateway-shaped benchmark rig that speaks structured System One, useful when you need apples-to-apples latency stories.

by @iamMrDuncan

benchmarkseval

Details Demo

BenchmarksPython19

mini-jev

Mini-Jev: what a Jev-style typed-decision interface looks like on a frozen Qwen3-4B, read the option letter's logits instead of generating JSON.

benchmarks

Details