Wav2vec2-large-xls-r-300m Amharic Speech Recognition Model - Open-source, Free, and Accurate Speech Recognition

Home

Wav2vec2 Large Xls R 300m Amharic Demo Colab

Developed by DipsankarSinha

Amharic speech recognition model fine-tuned on the common_voice_16_1 dataset based on facebook/wav2vec2-xls-r-300m

Speech Recognition

Transformers

Open Source License:Apache-2.0 #Amharic speech recognition #Low-resource language ASR #wav2vec2 fine-tuning

Downloads 18

Release Time : 6/23/2024

Model Overview

This is an automatic speech recognition (ASR) model optimized for Amharic, based on the wav2vec2-xls-r architecture and fine-tuned on the Common Voice dataset.

Model Features

Based on wav2vec2-xls-r architecture

Utilizes the powerful wav2vec2-xls-r-300m as the base model, with excellent speech feature extraction capabilities

Optimized for Amharic

Specially fine-tuned for Amharic to adapt to the phonetic characteristics of the language

Trained on Common Voice dataset

Trained using the Common Voice 16.1 dataset to ensure data diversity and representativeness

Model Capabilities

Amharic speech recognition

Audio to text conversion

Use Cases

Speech transcription

Amharic speech transcription

Convert Amharic speech to text

Word error rate 0.8639

Voice assistants

Amharic voice interaction

Used to build voice assistant applications supporting Amharic

Training Loss	Epoch	Step	Validation Loss	Wer
12.6948	5.0	100	4.1621	1.0
4.1026	10.0	200	4.0365	1.0
4.0037	15.0	300	3.9726	1.0007
3.9485	20.0	400	3.9524	1.0007
3.4635	25.0	500	2.4384	0.9980
1.1709	30.0	600	1.6987	0.9453
0.4955	35.0	700	1.5927	0.9073
0.3163	40.0	800	1.6750	0.8833
0.2372	45.0	900	1.6683	0.8813
0.1896	50.0	1000	1.6555	0.8779
0.1619	55.0	1100	1.6312	0.8819
0.1473	60.0	1200	1.6333	0.8639

Property	Details
Model Type	Fine - tuned version of facebook/wav2vec2-xls-r-300m
Training Data	common_voice_16_1

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 Xls R 300m Amharic Demo Colab

Model Overview

Model Features

Model Capabilities

Use Cases

🚀 wav2vec2-large-xls-r-300m-amharic-demo-colab

🚀 Quick Start

📚 Documentation

Model description

Intended uses & limitations

Training and evaluation data

🔧 Technical Details

Training procedure

Training hyperparameters

Training results

Framework versions

📄 License