Open-source xlsr-fa-lm speech recognition model - Fine-tuned based on Persian data for accurate speech recognition

Home

Xlsr Fa Lm

Developed by manifoldix

XLS-R-300m speech recognition model fine-tuned on Common Voice Persian data

Speech Recognition

Transformers

Other#Persian speech recognition #Low word error rate #Universal speech adaptation

Downloads 16

Release Time : 3/2/2022

Model Overview

This is an automatic speech recognition (ASR) model optimized for Persian, based on the XLS-R-300m architecture and fine-tuned on the Common Voice Persian dataset.

Model Features

Persian optimization

Specially optimized and fine-tuned for Persian speech recognition

XLS-R architecture

Based on the powerful XLS-R-300m architecture with excellent speech representation learning capabilities

Language model support

Supports integration with language models to improve recognition accuracy

Model Capabilities

Persian speech recognition

Speech-to-text conversion

Supports recognition without language model

Supports recognition with language model

Use Cases

Speech transcription

Persian speech transcription

Convert Persian speech content into text

Word error rate (WER) of 23-26%

Voice assistants

Persian voice command recognition

Used for voice command recognition in Persian voice assistants or control systems

Property	Details
Model Type	XLS-R-300m Wav2Vec2 Persian
Training Data	Common Voice

Task	Dataset	Metrics	Value
Speech Recognition (automatic - speech - recognition)	Common Voice fa	Test WER without LM (wer)	26%
Speech Recognition (automatic - speech - recognition)	Common Voice fa	Test WER with LM (wer)	23%

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

Xlsr Fa Lm

Model Overview

Model Features

Model Capabilities

Use Cases

🚀 XLSR-300m Persian

🚀 Quick Start

Model Information

Results

Demo Widgets

Tags