Kyle-han/roberta-base-nli-mean-tokens

Model Card

This repository contains a Jittor implementation and checkpoint for a Sentence-BERT style model.

Model Details

• Base model: roberta-base
• Pooling: mean
• Training data: NLI (SNLI + MultiNLI), with STS-B validation

Usage

This repository is intended for use with the Jittor SBERT code in this project.

Notes

• This is not an official release from the SBERT authors.