river-decider

Mobile ONNX encoder for the on-device agent of River (MonkeyMesh). It is a quantized export of fastino/GLiNER2.5-multi-Decide (Apache-2.0), derived from the community ONNX conversion DanKau/gliner2.5-multi-decide-onnx.

encoder.onnx: int8 embedding table, weights stored fp16 and computed in fp32. 364 MB. sha256 3078c2351e140f0c4f6f7a6d0349c167dd74f289c425e98490e9e807a0b5ec2a.
Max probability difference to the fp32 reference on 119 test posts: 0.054; labels 474/476 identical.
Use with onnx/classifier.onnx and tokenizer.json from the DanKau repo.

License: Apache-2.0, same as the original model. All credit for the model goes to Fastino.

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for adiZu/river-decider

Base model

Finetuned

Quantized

(4)

this model