devwithjevExtracts Clinical Variables from Notes
MedJev uses a 0.8B model and a 43 MB LoRA adapter trained to extract 11 predefined clinical variables from clinical notes.
JunMa_AI4Health@JunMa_AI4Healthπ
Curating clinical variables from free-text notes is tedious. General LLMs can help, but processing thousands of notes can be slow and costly. Inspired by Jev and the open-source community, weβre releasing MedJev to turn clinical notes into structured fields on consumer GPUs. A 0.8B model + 43 MB LoRA adapter, trained to extract 11 predefined clinical variables. On our benchmark of 2,895 held-out n
It turns notes into structured fields for use on consumer GPUs. The authors say processing thousands of notes with general LLMs can be slow and costly.
Also filed under Research & data
- GraphRAG with Swappable Laya and Jev Models
This project is an agentic GraphRAG pipeline with swappable local Laya or cloud Jev decision models.
- Jev Structured Decision Model Demo
Jev is TypeSafe AI's structured decision model. The repository contains research notes and an interactive Cloudflare Worker demo for it.
- Jev Question Workflows for Python
Hunch is a Python library that maps Jev Choice, Score, and Noul questions to classification, ranking, extraction, verification, and DataFrame workflows.
- Checks AI Visibility for YC Companies
Jev checked company names across 6,030 ChatGPT answers to buyer questions about 1,005 YC companies, producing 37,636 verdicts. The author reports that it took 40 seconds and cost $0.38.