Skip to content

Dataset · ehabnegm

100 hours of spoken Egyptian Arabic (single speaker)

ehabnegm/100-hour-Egyptian-dataset-single-speaker

Spoken Egyptian Arabic dialect audio corpus with transcripts for TTS and ASR training.

What's inside

The dataset contains 100 hours of live spoken speech in the Egyptian dialect of Arabic (arz/ar), read by a single speaker. Inside are 15,653 audio clips (24 kHz, mono) with aligned text transcripts. The authors compiled this corpus to address the shortage of open data for the spoken dialect, since most available databases are recorded in Modern Standard Arabic.

What you can do with it

  • Training text-to-speech (TTS) models with natural Egyptian pronunciation
  • Fine-tuning automatic speech recognition (ASR) systems to understand the spoken Arabic dialect
Tasks
Text → speech, Automatic speech recognition
Languages
Arabic, Egyptian Arabic
Size
10K–100K rows
Downloads this month
157.6K
Likes
15

How to load it

from datasets import load_dataset

ds = load_dataset("ehabnegm/100-hour-Egyptian-dataset-single-speaker")