wav2vec2-xls-r-1b-arabic Open-source Model - Supports Precise Arabic Speech-to-Text Conversion

Wav2vec2 Xls R 1b Arabic

Developed by AndrewMcDowell

This model is an automatic speech recognition (ASR) model fine-tuned on the Mozilla Common Voice 8.0 Arabic dataset based on facebook/wav2vec2-xls-r-1b, supporting Arabic speech-to-text tasks.

Speech Recognition

Transformers

ArabicOpen Source License:Apache-2.0 #Arabic speech recognition #Large model fine-tuning #Common Voice dataset

Downloads 20

Release Time : 3/2/2022

Model Overview

This is an optimized automatic speech recognition (ASR) model for Arabic, capable of converting Arabic speech into text. The model achieved a word error rate (WER) of 0.8607 on the Common Voice 8.0 Arabic evaluation set.

Model Features

Large-scale pre-training

Fine-tuned on the 1-billion-parameter wav2vec2-xls-r-1b model, featuring powerful speech feature extraction capabilities.

Arabic optimization

Specifically optimized for Arabic speech characteristics, performing well on the Common Voice Arabic dataset.

Open-source license

Licensed under Apache 2.0, allowing for both commercial and research use.

Model Capabilities

Arabic speech recognition

Speech-to-text

Real-time speech processing

Use Cases

Speech transcription

Arabic speech transcription

Convert Arabic speech content into text format

Word error rate 0.8607

Voice assistants

Arabic voice command recognition

Used for Arabic voice assistants and smart home control

🚀 Automatic Speech Recognition Model

This model is a fine - tuned version of facebook/wav2vec2-xls-r-1b on the MOZILLA - FOUNDATION/COMMON_VOICE_8_0 - AR dataset. It offers high - performance automatic speech recognition capabilities.

🚀 Quick Start

This model is ready to use for automatic speech recognition tasks. You can fine - tune it further based on your specific requirements.

📚 Documentation

Model Performance

It achieves the following results on the evaluation set:

Loss: 1.1373
Wer: 0.8607

Training and Evaluation Data

The model is trained on the MOZILLA - FOUNDATION/COMMON_VOICE_8_0 - AR dataset. However, more detailed information about the training and evaluation data is yet to be provided.

Training Procedure

Training Hyperparameters

The following hyperparameters were used during training:

learning_rate: 6.5e - 05
train_batch_size: 16
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 64
optimizer: Adam with betas=(0.9,0.999) and epsilon = 1e - 08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 2000
num_epochs: 30.0
mixed_precision_training: Native AMP

Training Results

Training Loss	Epoch	Step	Validation Loss	Wer
2.2416	0.84	500	1.2867	0.8875
2.3089	1.67	1000	1.8336	0.9548
2.3614	2.51	1500	1.5937	0.9469
2.5234	3.35	2000	1.9765	0.9867
2.5373	4.19	2500	1.9062	0.9916
2.5703	5.03	3000	1.9772	0.9915
2.4656	5.86	3500	1.8083	0.9829
2.4339	6.7	4000	1.7548	0.9752
2.344	7.54	4500	1.6146	0.9638
2.2677	8.38	5000	1.5105	0.9499
2.2074	9.21	5500	1.4191	0.9357
2.3768	10.05	6000	1.6663	0.9665
2.3804	10.89	6500	1.6571	0.9720
2.3237	11.72	7000	1.6049	0.9637
2.317	12.56	7500	1.5875	0.9655
2.2988	13.4	8000	1.5357	0.9603
2.2906	14.24	8500	1.5637	0.9592
2.2848	15.08	9000	1.5326	0.9537
2.2381	15.91	9500	1.5631	0.9508
2.2072	16.75	10000	1.4565	0.9395
2.197	17.59	10500	1.4304	0.9406
2.198	18.43	11000	1.4230	0.9382
2.1668	19.26	11500	1.3998	0.9315
2.1498	20.1	12000	1.3920	0.9258
2.1244	20.94	12500	1.3584	0.9153
2.0953	21.78	13000	1.3274	0.9054
2.0762	22.61	13500	1.2933	0.9073
2.0587	23.45	14000	1.2516	0.8944
2.0363	24.29	14500	1.2214	0.8902
2.0302	25.13	15000	1.2087	0.8871
2.0071	25.96	15500	1.1953	0.8786
1.9882	26.8	16000	1.1738	0.8712
1.9772	27.64	16500	1.1647	0.8672
1.9585	28.48	17000	1.1459	0.8635
1.944	29.31	17500	1.1414	0.8616

Framework Versions

Transformers 4.17.0.dev0
Pytorch 1.10.2+cu102
Datasets 1.18.2.dev0
Tokenizers 0.11.0

📄 License

This model is released under the Apache - 2.0 license.

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