by rorshopping · AI Tool · ★ 25
Jev-style decisions on a laptop Unofficial research repo. Not affiliated with TypeSafe AI. This is a hands-on study of the parallel constrained decoding idea behind Jev — the "System One" model launched by Diogo Almeida's TypeSafe AI in Sep 2026 — reproduced on a stock, untrained small model, on an Apple Silicon laptop. The trick: instead of generating JSON token-by-token, define your schema as a set of typed fields, prefill the context once, broadcast the KV cache across one batch row per field, and evaluate all fields in a single forward pass.
| Stars | 25 |
| Forks | 1 |
| Language | Python |
| Category | AI Tool |
| Quality Score | 58.8891041860101/100 |
| Open Issues | 1 |
| Last Updated | 2026-09-17 |
| Created | 2026-09-16 |
| Platforms | python |
| Est. Tokens | ~15k |
These tools work well together with jev-on-a-laptop for enhanced workflows:
Explore other popular ai tool tools:
jev-on-a-laptop is Unofficial study: Jev-style parallel typed decisions on stock 1.5B-8B models on an Apple Silicon laptop. Benchmarks, research notes, and a Hugging Face Space demo.. It is categorized as a AI Tool with 25 GitHub stars.
jev-on-a-laptop is primarily written in Python.
You can find installation instructions and usage details in the jev-on-a-laptop GitHub repository at github.com/rorshopping/jev-on-a-laptop. The project has 25 stars and 1 forks, indicating an active community.