Hanhpt23/whisper-tiny-silvar

Hanhpt23/whisper-tiny-slisa

This model is a fine-tuned version of Hanhpt23/whisper-tiny-slisa on the Hanhpt23/Silvar dataset. It achieves the following results on the evaluation set:

• Loss: 0.1569
• Wer: 5.4567

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: 8
• 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: 100
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0875	1.0	113	0.0998	3.5843
0.0458	2.0	226	0.1395	5.3098
0.0142	3.0	339	0.1570	6.6358
0.0182	4.0	452	0.1750	6.5465
0.0213	5.0	565	0.1586	6.1090
0.0027	6.0	678	0.1691	5.8729
0.0006	7.0	791	0.1623	5.7367
0.0004	8.0	904	0.1578	5.1751
0.0003	9.0	1017	0.1570	5.4733
0.0002	10.0	1130	0.1569	5.4567

Framework versions

• Transformers 4.41.1
• Pytorch 2.3.0
• Datasets 2.19.1
• Tokenizers 0.19.1