AIbase
EN
Home
Wav2vec2 Large Xls R 300m Greek
W

Wav2vec2 Large Xls R 300m Greek

Developed by jerrychatz
Greek speech recognition model trained on the Common Voice dataset, using the wav2vec2 architecture
Speech Recognition
Transformers
#Greek speech recognition#Large parameter speech model#Low word error rate
Downloads 19
Release Time : 3/2/2022

Model Overview

This model is an automatic speech recognition (ASR) model for Greek, trained on Facebook's wav2vec2 architecture, suitable for converting Greek speech to text.

Model Features

Large-scale pretraining
Based on a large-scale model architecture with 300M parameters, featuring powerful speech feature extraction capabilities
Greek language optimization
Specifically trained and optimized for Greek speech data
Efficient training
Utilizes mixed-precision training and gradient accumulation techniques to improve training efficiency

Model Capabilities

Greek speech recognition
Speech-to-text
Continuous speech recognition

Use Cases

Speech transcription
Greek meeting minutes
Automatically convert Greek meeting recordings into text transcripts
Word error rate approximately 33.38%
Voice assistant
Provide speech recognition capabilities for Greek voice assistants
Education
Language learning applications
Assist Greek language learners in practicing pronunciation and listening
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
EnglishįŽ€äŊ“中文įšéĢ”中文ãĢãģんご
Š 2025AIbase