Ascend_with_timit Open-source Speech Recognition Model - Free Deployment for Accurate Speech-to-Text Conversion

Ascend With Timit

Developed by GleamEyeBeast

This model is a speech recognition model fine-tuned on the TIMIT dataset, achieving a word error rate of 0.4781 and a character error rate of 0.1727 on the evaluation set.

Speech Recognition

Transformers

#Speech Recognition #Low Word Error Rate #TIMIT Fine-tuning

Downloads 16

Release Time : 4/4/2022

Model Overview

This is an Automatic Speech Recognition (ASR) model primarily used to convert speech into text. The model is fine-tuned on the TIMIT dataset and is suitable for English speech recognition tasks.

Model Features

Low Word Error Rate

Achieved a word error rate of 0.4781 on the evaluation set, demonstrating good performance.

Low Character Error Rate

Achieved a character error rate of 0.1727 on the evaluation set, showing high accuracy.

Efficient Training

Optimized training efficiency using mixed-precision training (native AMP).

Model Capabilities

English Speech Recognition

Speech-to-Text

Use Cases

Speech Transcription

Meeting Minutes

Automatically convert meeting recordings into text transcripts

Accuracy approximately 52.19% (based on 1-WER calculation)

Subtitle Generation

Automatically generate English subtitles for video content

Character-level accuracy approximately 82.73%

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
2.4026	1.0	890	1.3419	0.9083	0.3670
1.1926	2.0	1780	0.9730	0.6491	0.2585
0.9104	3.0	2670	0.8483	0.5368	0.1963
0.7718	4.0	3560	0.8122	0.4913	0.1791
0.7013	5.0	4450	0.8013	0.4781	0.1727

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

Ascend With Timit

Model Overview

Model Features

Model Capabilities

Use Cases

🚀 ascend_with_timit

🚀 Quick Start

📚 Documentation

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions