W

Wav2vec2 Large Xlsr 53 Greek

Developed by jonatasgrosman
This is a large XLSR-53 model fine-tuned for Greek speech recognition tasks, based on facebook/wav2vec2-large-xlsr-53 model, trained using Common Voice 6.1 and CSS10 datasets.
Downloads 130.81k
Release Time : 3/2/2022

Model Overview

This model is specifically designed for Greek automatic speech recognition (ASR), capable of converting Greek speech into text.

Model Features

High-performance Greek recognition
Achieves 11.62% word error rate (WER) and 3.36% character error rate (CER) on Common Voice Greek test set
Based on XLSR-53 large model
Fine-tuned from facebook/wav2vec2-large-xlsr-53 model with powerful speech feature extraction capabilities
Multi-dataset training
Trained using Common Voice 6.1 and CSS10 datasets, covering diverse speech scenarios

Model Capabilities

Greek speech recognition
16kHz audio processing
Direct usage without language model

Use Cases

Speech-to-text
Greek speech transcription
Convert Greek speech content into text
Accuracy reaches 88.38% (1-WER)
Voice assistants
Greek voice command recognition
Used for Greek voice assistant or control system command recognition
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase