AIbase
EN
Home
Wav2vec2 Large Xls R 300m Ru
W

Wav2vec2 Large Xls R 300m Ru

Developed by mobedkova
This is a Russian automatic speech recognition model based on the Wav2Vec2 XLS-R architecture with a parameter scale of 300m, evaluated on public speech and robust speech event datasets.
Speech Recognition
Transformers
 Other#Russian speech recognition#Multi-scenario robustness#XLS-R architecture
Downloads 37
Release Time : 3/2/2022

Model Overview

This model is primarily used for Russian speech recognition tasks, capable of converting Russian speech into text.

Model Features

High-performance Russian speech recognition
Achieved a word error rate of 27.81% and a character error rate of 8.83% on the Common Voice-7.0 Russian dataset.
Robust performance
Performed well on the Robust Speech Event dataset, with word error rates of 44.64% and 42.51% for development and test data, respectively.
Based on Wav2Vec2 XLS-R architecture
Utilizes the advanced Wav2Vec2 XLS-R architecture with powerful speech feature extraction capabilities.

Model Capabilities

Russian speech recognition
Speech-to-text

Use Cases

Speech transcription
Russian meeting minutes
Automatically transcribe Russian meeting recordings into text records
Word error rate 27.81% (Common Voice dataset)
Russian voice assistant
Speech recognition module for Russian voice assistants
Speech analysis
Russian speech content analysis
Analyze Russian speech content to extract key information
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
English简体中文繁體中文にほんご
© 2025AIbase