wav2vec2-large-960h-lv60-self-4-gram_fine-tune_real_29_Jun Open-source Speech Recognition Model

Wav2vec2 Large 960h Lv60 Self 4 Gram Fine Tune Real 29 Jun

Developed by RuiqianLi

This model is a fine-tuned speech recognition model based on facebook/wav2vec2-large-960h-lv60-self on the uob_singlish dataset, supporting Singapore English recognition.

Speech Recognition

Transformers

Open Source License:Apache-2.0 #Singapore English Recognition #Speech to Text #Low Resource Optimization

Downloads 105

Release Time : 6/29/2022

Model Overview

This is a fine-tuned speech recognition model specifically optimized for Singapore English (uob_singlish dataset), achieving a word error rate of 0.4583 on the evaluation set.

Model Features

Singapore English Optimization

Specifically fine-tuned for Singapore English (uob_singlish dataset)

Low Word Error Rate

Achieved a word error rate of 0.4583 on the evaluation set

Based on wav2vec2 Architecture

Uses facebook's wav2vec2-large-960h-lv60-self as the base model

Model Capabilities

Speech Recognition

Singapore English Recognition

Use Cases

Speech to Text

Singapore English Speech Transcription

Convert Singapore English speech content into text

Word error rate 0.4583

Training Loss	Epoch	Step	Validation Loss	Wer
2.1283	1.82	20	1.5236	0.5764
1.3015	3.64	40	1.2956	0.4931
0.9918	5.45	60	1.3087	0.5347
0.849	7.27	80	1.2914	0.5139
0.6191	9.09	100	1.2895	0.4583

Property	Details
Model Type	Fine - tuned wav2vec2-large-960h-lv60-self
Training Data	uob_singlish

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご

Wav2vec2 Large 960h Lv60 Self 4 Gram Fine Tune Real 29 Jun

Model Overview

Model Features

Model Capabilities

Use Cases

🚀 wav2vec2-large-960h-lv60-self-4-gram_fine-tune_real_29_Jun

🚀 Quick Start

📚 Documentation

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

📄 License