Byne/whisper-large-v3-arabic

arabic

This model is a fine-tuned version of openai/whisper-large-v3 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1384
• Wer: 9.3775

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: 1e-05
• train_batch_size: 32
• eval_batch_size: 16
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 4
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.1656	0.4963	875	0.1865	14.2234
0.1473	0.9926	1750	0.1515	14.5240
0.0949	1.4889	2625	0.1441	11.1473
0.0762	1.9853	3500	0.1347	12.8537
0.0459	2.4816	4375	0.1380	11.4366
0.0433	2.9779	5250	0.1312	10.1302
0.0106	3.4742	6125	0.1406	9.6804
0.0124	3.9705	7000	0.1384	9.3775

Framework versions

• Transformers 4.51.3
• Pytorch 2.1.0+cu118
• Datasets 3.5.0
• Tokenizers 0.21.1