jevhome/jevhome-E

jevhome E

A typed-decision model (bi-encoder, ~48M parameters, fp32) for jevhome: give it a state and yes/no (noul), pick-one (choice) or graded (score) questions, and it returns calibrated probabilities in one forward pass, on a plain CPU. Same API as Jev (POST /v1/systemone).

The order of the options of a choice question does not change the answer: each option is embedded on its own.

pytorch/ holds the PyTorch weights (backbone model.safetensors + decision head head.safetensors) for GPU or research use; jevhome only needs the ONNX files (--exclude 'pytorch/*' skips them).

Run it

git clone https://github.com/Jibril-Frej/jev-at-home && cd jev-at-home && cargo build --release    # the jevhome binary (needs Rust)
hf download jevhome/jevhome-E --local-dir models/jevhome-E --exclude 'pytorch/*'
./target/release/jevhome serve models/jevhome-E

Numbers

JevBench Easy	Standard	Hard	ECE (Easy+Standard, lower is better)	External mean (10 sets)	Latency p50	p90	Peak RAM
88	64	30	0.125	55.0	37 ms	88 ms	0.3 GB

Accuracy in %, measured with the jevhome binary. Latency: one decision at a time, 4 threads on 4 cores of an AMD EPYC 9654. The full tables (every set, Jev, Laya and the teacher), the training recipe and the data are in the jevhome README.

Files

• encoder.onnx, scorer.onnx: the network, probabilities come from calibrated logits.
• decision_config.json, tokenizer*.json, config.json: input format and tokenizer.
• calibration_u.json: one temperature per question type and number of options.

Licence

CC BY-NC 4.0 (non-commercial): several training sources are non-commercial or share-alike. Backbone: ibm-granite/granite-embedding-small-english-r2.