This benchmark contains Quranic recitation audio from three sources, including REAL USER PHONE RECORDINGS (Tarteel tlog) and studio recitations, redistributed here only to enable reproducible ASR evaluation. By requesting access you agree: (1) to use it solely for ASR research / evaluation, (2) NOT to redistribute the audio or use it for voice cloning / TTS / re-identification, and (3) to respect the reciters' and users' rights and the upstream dataset terms (tarteel-ai/tlog, tarteel-ai/everyayah, QUL).

Quranic ASR Benchmark - leakage-free, held-out

A small, leakage-free benchmark (600 clips) for evaluating Arabic ASR on Quranic recitation (Hafs riwayah). Every clip is verified absent from our training data, so it measures generalization, not memorization. Same clips + same scoring for every model.

📊 Live leaderboard: https://huggingface.co/spaces/Muno459/quranic-asr-leaderboard

The set (600 clips, 200 per source)

Source	n	What it is
`everyayah_heldout`	200	3 EveryAyah reciters with zero training samples (clean studio).
`qul_alnufais`	200	A QUL reciter (Al-Nufais) never trained on; references from the canonical mushaf.
`tlog_holdout`	200	Real phone-mic recitation held out from training (the hard, real-world case).

tlog is the one that matters for apps; studio numbers flatter every model.

Hugging Face loading

Each source is exposed as its own dataset configuration with a single test split. Rows include audio, text, id, and source, so ASR evaluation can be run without extra reference-transcript preprocessing.

from datasets import load_dataset

ds = load_dataset("Muno459/quranic-asr-benchmark", "tlog_holdout", split="test")
print(ds[0]["audio"], ds[0]["text"])

Available configs:

everyayah_heldout - clean studio held-out EveryAyah reciters
qul_alnufais - held-out QUL Al-Nufais reciter
tlog_holdout - held-out real phone-mic Tarteel tlog recordings

Files

benchmark.jsonl - official scoring manifest: id, source, reference_text, audio (relative path).
audio/<source>/metadata.jsonl - Hugging Face AudioFolder metadata exposing text for each clip.
audio/<source>/*.wav - 16 kHz mono audio.
score.py - the official scorer (exact normalization + WER/CER, per-source + overall).
results.json - current leaderboard results.

How to evaluate your model

# 1. transcribe every clip in benchmark.jsonl (use the `audio` path), write one hyp per `id`:
#    {"id": "everyayah_heldout/xxx", "text": "..."}   (jsonl)  OR  {id: text} (json)
# 2. score (same normalization as the leaderboard):
pip install rapidfuzz
python score.py --hyps my_hyps.jsonl

Scoring / normalization

WER & CER over normalized text: combining marks + Quranic annotation signs + tatweel stripped; hamza forms and alef-maqsura unified. The alef-insensitive column additionally strips alef to neutralize the Uthmani-rasm vs imlaei spelling-convention mismatch (eval refs are rasm-ish; many models output imlaei), reflecting true word accuracy over orthography. CTC/greedy or each model's default decoder; Arabic forced/auto-detected per model.

Leakage control

The public EveryAyah test split overlaps most Quran ASR training sets, so we use only the EveryAyah reciters with zero training overlap, plus a held-out QUL reciter and held-out real-phone tlog. nvidia FastConformer (current public #1) is scored on the same clips for a fair head-to-head.

Acknowledgements

Special thanks to David Jiménez Paredes / deepdml for the original community effort around the Open Universal Arabic Quranic ASR Leaderboard, which helped inspire this benchmark and leaderboard work. His feedback also directly improved this dataset's Hugging Face structure by encouraging source-specific configurations and direct transcript fields for easier downstream evaluation.

Provenance & licensing

Audio is redistributed for research reproducibility under a manual-approval gate. Upstream sources and their terms: EveryAyah (tarteel-ai/everyayah, CC-BY 4.0), tlog (tarteel-ai/tlog, gated), QUL (qul.tarteel.ai). Respect each upstream license; do not redistribute the audio. Reciters and tlog users retain their rights. If you are a rights-holder and want a clip removed, open a discussion.

Downloads last month: 17

Total file size:

247 MB

You need to agree to share your contact information to access this dataset