The open-source speech recognition model wav2vec2-base-timit-demo - Efficiently achieve precise speech-to-text conversion

Wav2vec2 Base Timit Demo

Developed by dlu66061

A speech recognition model fine-tuned based on facebook/wav2vec2-base, achieving a 28.25% word error rate on the TIMIT dataset

Speech Recognition

Transformers

Open Source License:Apache-2.0 #Speech Recognition #Low Word Error Rate #TIMIT Dataset

Downloads 21

Release Time : 4/20/2022

Model Overview

This is a pre-trained model for English speech recognition, fine-tuned based on the wav2vec2 architecture, suitable for Automatic Speech Recognition (ASR) tasks

Model Features

Low Word Error Rate

Achieves a 28.25% word error rate (WER) on the evaluation set

Based on wav2vec2 Architecture

Uses facebook's wav2vec2-base as the base model

End-to-End Training

Learns speech representations directly from raw audio without manual feature extraction

Model Capabilities

English Speech Recognition

Audio to Text

Automatic Speech Transcription

Use Cases

Speech Transcription

Meeting Minutes

Automatically convert meeting recordings into text transcripts

Accuracy approximately 71.75%

Voice Notes

Convert voice memos into searchable text

Training Loss	Epoch	Step	Validation Loss	Wer
3.5419	3.45	500	1.2376	0.8772
0.5393	6.9	1000	0.4489	0.3894
0.1916	10.34	1500	0.3777	0.3185
0.1139	13.79	2000	0.4041	0.3058
0.0798	17.24	2500	0.3742	0.2988
0.0602	20.69	3000	0.3751	0.2897
0.0463	24.14	3500	0.4067	0.2865
0.0388	27.59	4000	0.4094	0.2825

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 Base Timit Demo

Model Overview

Model Features

Model Capabilities

Use Cases

🚀 wav2vec2-base-timit-demo

🚀 Quick Start

📚 Documentation

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

📄 License