Datasets:
The dataset viewer is not available for this split.
Parquet error: Scan size limit exceeded: attempted to read 1251276181 bytes, limit is 300000000 bytes
Make sure that
1. the Parquet files contain a page index to enable random access without loading entire row groups2. otherwise use smaller row-group sizes when serializing the Parquet files
Error code: TooBigContentError
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
everyayah-wav — Quranic recitation audio mirror
Full-mushaf Quranic recitation audio at 16 kHz mono 16-bit WAV, re-encoded from everyayah.com for ML / ASR research.
This dataset is intentionally audio-only — no transcription text and no alignment timings. The canonical Quranic text is widely available from Tanzil and other public sources; pair this audio with whatever text edition fits your use case.
Schema
| Column | Type | Notes |
|---|---|---|
audio |
Audio(16000) |
16 kHz mono 16-bit WAV bytes |
duration |
float32 |
seconds |
reciter |
string |
one of ~44 reciters (config name) |
surah |
int32 |
1..114 |
ayah |
int32 |
1-indexed within surah |
Usage
from datasets import load_dataset
ds = load_dataset('dev-ahmedhany/everyayah-wav', name='Maher_AlMuaiqly', split='train')
import io, soundfile as sf
row = ds[0]
audio_array, sr = sf.read(io.BytesIO(row['audio']['bytes']))
print(row['surah'], row['ayah'], audio_array.shape, sr)
Source / license
Audio re-encoded from public-domain everyayah.com MP3s. CC0 1.0.
Built for mehrab.
- Downloads last month
- 225