Model reference · open weights

whisper-tiny-ar-quran

Audio tarteel-ai Speech→text 1 build Open weights 1k dl/mo

whisper-tiny-ar-quran is an open-weight audio or speech model from tarteel-ai. whisper-tiny-ar-quran (BF16) weighs 151 MB; the smallest configuration that runs it is RTX 3060 12 GB.

Request Rent a GPU

For self-hosting, compare this build's requirements below with the GPU you plan to rent.

whisper-tiny-ar-quran is an automatic speech recognition model developed by tarteel-ai.
It is a fine-tuned version of openai/whisper-tiny and is licensed under apache-2.0.
The model achieves a word error rate of 7.0535 on its evaluation set.

Summary of the tarteel-ai/whisper-tiny-ar-quran model card, 2026-10-01

Hugging Face ↗

What it is

Released by	tarteel-ai
Released	2022-12-08
VRAM	151 MB for the weights

What it runs on

Memory and cards for whisper-tiny-ar-quran (BF16)

151 MBweights, file size

1.6 GBruntime overhead

Card	One stream	Memory
RTX 3060 12 GB	fits	11.6 GB
RTX 4060 Ti 16 GB	fits	15.4 GB
RTX 3090 24 GB	fits	23.4 GB
RTX 4090 24 GB	fits	23.4 GB
RTX 5090 32 GB	fits	31.0 GB
L40S 48 GB	fits	44.0 GB
A100 80 GB	fits	78.2 GB
H100 80 GB	fits	78.1 GB
RTX PRO 6000 Blackwell 96 GB	fits	93.8 GB
DGX Spark (GB10) 128 GB unified	fits	107 GB
H200 141 GB	fits	138 GB
B200 180 GB	fits	176 GB

From the model card

What tarteel-ai says about whisper-tiny-ar-quran

Read the model card

This model is a fine-tuned version of openai/whisper-tiny on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.0928
Wer: 7.0535

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0001
train_batch_size: 16
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 500
training_steps: 5000
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.1766	0.05	500	0.2829	20.0236
0.1129	0.09	1000	0.1981	13.8364
0.0775	0.14	1500	0.1763	12.5450
0.0678	0.19	2000	0.1485	10.7302
0.0437	0.23	2500	0.1336	9.6693
0.0341	0.28	3000	0.1244	8.9602
0.0302	0.33	3500	0.1059	8.2224
0.0189	0.37	4000	0.1044	7.6902
0.0167	0.42	4500	0.0966	7.2643
0.0151	0.47	5000	0.0928	7.0535

Framework versions

Transformers 4.26.0.dev0
Pytorch 1.13.0+cu117
Datasets 2.7.1.dev0
Tokenizers 0.13.2

Quoted from the model card on Hugging Face. The full card is behind the Hugging Face link above.

Details

Type	Audio & music
Task	Speech→text
Runs with	transformers
Popularity	1k downloads / month
Weights	151 MB, whisper-tiny-ar-quran (BF16, file size)