AIbase
EN
Home
Wav2vec2 Large Xls R 300m Welsh
W

Wav2vec2 Large Xls R 300m Welsh

Developed by infinitejoy
This is an automatic speech recognition model fine-tuned on the Welsh dataset based on facebook/wav2vec2-xls-r-300m, achieving a word error rate of 31.003% and a character error rate of 7.775% on the Common Voice 7 Welsh test set.
Speech Recognition
Transformers
 OtherOpen Source License:Apache-2.0#Welsh speech recognition#Multilingual speech processing#Low word error rate
Downloads 89
Release Time : 3/2/2022

Model Overview

This model is an automatic speech recognition (ASR) model optimized for Welsh, capable of converting Welsh speech into text.

Model Features

Welsh optimization
Specifically fine-tuned for Welsh speech recognition, performing well on the Common Voice 7 Welsh test set
Based on XLS-R architecture
Utilizes facebook's wav2vec2-xls-r-300m pre-trained model as the foundation, with strong speech feature extraction capabilities
Low character error rate
Achieved a character error rate (CER) of 7.775% on the test set, indicating high accuracy in recognizing individual characters

Model Capabilities

Welsh speech recognition
Speech-to-text
Conversational speech processing

Use Cases

Speech transcription
Welsh speech transcription
Convert Welsh speech content into text format
Character error rate 7.775%, word error rate 31.003%
Voice assistants
Welsh voice command recognition
Used to support Welsh-language voice assistants and smart devices
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