Mandarin_naive Open-source Mandarin Speech Recognition Model - Precise Recognition for Voice Interaction Applications

Mandarin Naive

Developed by GleamEyeBeast

This model is a Mandarin speech recognition model fine-tuned on the Common Voice dataset based on facebook/wav2vec2-xls-r-300m

Speech Recognition

Transformers

Open Source License:Apache-2.0 #Mandarin speech recognition #Fine-tuned XLS-R #Low word error rate

Downloads 22

Release Time : 3/2/2022

Model Overview

A speech recognition model optimized for Mandarin, fine-tuned on the wav2vec2-xls-r-300m architecture and trained on the Common Voice dataset

Model Features

Mandarin optimization

Specially fine-tuned for Mandarin speech recognition tasks

Based on wav2vec2 architecture

Uses facebook's wav2vec2-xls-r-300m pre-trained model as the foundation

Medium scale

300M parameter model size, balancing performance and computational resource requirements

Model Capabilities

Mandarin speech recognition

Speech-to-text

Use Cases

Speech transcription

Mandarin speech transcription

Convert Mandarin speech content into text

Word Error Rate (WER) 0.3999

Voice assistants

Mandarin voice command recognition

Used for command recognition in voice assistant systems

Training Loss	Epoch	Step	Validation Loss	Wer
4.8963	3.67	400	1.0645	0.8783
0.5506	7.34	800	0.5032	0.5389
0.2111	11.01	1200	0.4765	0.4712
0.1336	14.68	1600	0.4815	0.4511
0.0974	18.35	2000	0.4956	0.4370
0.0748	22.02	2400	0.4881	0.4235
0.0584	25.69	2800	0.4732	0.4193
0.0458	29.36	3200	0.4584	0.3999

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

Mandarin Naive

Model Overview

Model Features

Model Capabilities

Use Cases

🚀 Mandarin_naive

🚀 Quick Start

📚 Documentation

Model description

Intended uses & limitations

Training and evaluation data

🔧 Technical Details

Training procedure

Training hyperparameters

Training results

Framework versions

📄 License