otozz
/

whisper-small-ar_tsize_0.8

Automatic Speech Recognition

Model card Files Files and versions

Details

Model trained with train_size=0.8 or 80% of the Mozilla Common Voice Arabic training+validation dataset.

Downloads last month: 1

Safetensors

Model size

0.2B params

Tensor type

F32

·

Collections including otozz/whisper-small-ar_tsize_0.8

finetune-ar-dialects

Models for the thesis titled: "The Effects of Fine-Tuning on the ASR Performance of Dialectal Arabic". • 17 items • Updated May 20, 2024 • 3

Overcoming Data Scarcity in Multi-Dialectal Arabic ASR

Models and datasets for the paper "Overcoming Data Scarcity in Multi-Dialectal Arabic ASR via Whisper Fine-Tuning". https://arxiv.org/abs/2506.02627 • 54 items • Updated Jun 5, 2025 • 1