W

Wav2vec2 Large Xls R 300m Bg D2

Developed by DrishtiSharma
An automatic speech recognition model fine-tuned on Bulgarian language datasets based on facebook/wav2vec2-xls-r-300m
Downloads 20
Release Time : 3/2/2022

Model Overview

This model is an automatic speech recognition (ASR) model for Bulgarian, fine-tuned on the Common Voice 8.0 dataset, capable of converting Bulgarian audio into text.

Model Features

High-performance Bulgarian recognition
Achieves a WER of 28.78% and CER of 6.86% on the Common Voice 8.0 test set, demonstrating excellent performance
Based on large-scale pre-trained models
Fine-tuned from facebook's wav2vec2-xls-r-300m model, inheriting powerful speech feature extraction capabilities
Multi-dataset validation
Validated not only on the Common Voice dataset but also on robust speech event datasets

Model Capabilities

Bulgarian audio to text conversion
Long audio processing (supports chunk processing)
Speech recognition evaluation

Use Cases

Speech transcription
Bulgarian speech to text
Convert Bulgarian speech content into editable text
High accuracy on standard test sets
Voice assistants
Bulgarian voice command recognition
Basic recognition module for building Bulgarian voice assistants
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase