R

Roberta Base Use Qa Bg

Developed by rmihaylov
This is a multilingual Roberta model that can generate sentence embeddings for Bulgarian, inspired by Sentence-BERT with Google's USE model as the teacher.
Downloads 14
Release Time : 4/18/2022

Model Overview

The model generates Bulgarian sentence embeddings by mapping translated sentences to the same vector space position as the original for sentence similarity calculation.

Model Features

Multilingual support
Supports generating sentence embeddings for Bulgarian and English.
Case-sensitive
The model is case-sensitive, e.g., 'bulgarian' and 'Bulgarian' are treated differently.
Based on Sentence-BERT concept
Uses the method of mapping translated sentences to the same vector space position.
Teacher model usage
Google's USE model was used as the teacher during training.

Model Capabilities

Generate sentence embeddings
Calculate sentence similarity
Support Bulgarian processing
Support English processing

Use Cases

QA systems
Question-answer matching
Calculate similarity between user questions and candidate answers
Examples demonstrate how to find the most relevant answers to questions.
Information retrieval
Relevant document retrieval
Find the most relevant documents based on query sentences
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase