zeinhasan/whisper-small-id

Whisper Tiny id - Zein

This model is a fine-tuned version of openai/whisper-tiny on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.9557
• Wer: 49.1561

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: 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: 6000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.3251	3.1646	1000	0.7617	48.4559
0.1132	6.3291	2000	0.7951	47.6630
0.0414	9.4937	3000	0.8504	48.5208
0.0146	12.6582	4000	0.9079	49.3647
0.0083	15.8228	5000	0.9421	48.9057
0.0064	18.9873	6000	0.9557	49.1561

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.0
• Datasets 3.0.0
• Tokenizers 0.19.1