lingura gec-universal-int4

Owned universal grammar-correction artifact for LINGURA (com.lingura.app).

INT8 ONNX build of EuroBERT/EuroBERT-210m at revision 39b51e15dd1f1a06f58b5cbf6a8a188cec60bd0e — EuroBERT encoder + Seq2Edit heads (action logits [B,T,9], sparse fill token logits, language head). See metadata.json for the tensor contract, calibration and per-file sha256/bytes.

EuroBERT-210m — Nicolas Boizard, Hippolyte Gisserot-Boukhlef, Duarte M. Alves, et al. (EuroBERT: Scaling Multilingual Encoders for European Languages), licensed under Apache License 2.0.

Pipeline: scripts/gec/ in the LINGURA repo — torch.onnx.export (opset 18) + quantize_dynamic (MatMul+Gather → per-channel INT8).

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