Downloads · 30 days
7
16% of all-time downloads
adkar-recite-wav2vec2-ar-ctc
boutchaz/adkar-recite-wav2vec2-ar-ctc
adkar-recite-wav2vec2-ar-ctc is a machine learning model from boutchaz. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
Arabic wav2vec2-CTC acoustic model exported to ExecuTorch (.pte, int8 dynamic quantization) for on-device Quran recitation checking in the Adkar app.
All-time downloads
44
Public
Repo size
634 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pte634 MB · 100%
At a glance
- License
- apache-2.0
- Access
- Public
- Created
- Jul 19, 2026
- Updated
- Jul 19, 2026
- SHA
- ac0ac82f
Base models
Model card
From the Hugging Face model README
adkar-recite-wav2vec2-ar-ctc
Arabic wav2vec2-CTC acoustic model exported to ExecuTorch (.pte, int8 dynamic quantization) for on-device Quran recitation checking in the Adkar app.
- Base model:
jonatasgrosman/wav2vec2-large-xlsr-53-arabic(Apache-2.0) - Input:
[1, T]float32 mono waveform @ 16 kHz,Ta multiple of 320 - Output:
[1, T/320, 52]character logits (vocab invocab_ar.json) - Usage: CTC forced alignment against a known target text (no open-vocab decoding) — see the app's
reciteCTCgrader. - Runtime:
react-native-executorchgenericExecutorchModule.
Not a general-purpose ASR release — weights are unchanged from the base model apart from export + quantization.
- License
- apache-2.0
- Languages
- ar
- Created
- Jul 19, 2026
- Updated
- Jul 19, 2026