whisper_attention_0010

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

• Train Loss: 4.5235
• Train Accuracy: 0.0121
• Train Wermet: 1.0483
• Validation Loss: 3.7736
• Validation Accuracy: 0.0118
• Validation Wermet: 1.4279
• Epoch: 9

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:

• optimizer: {'name': 'AdamWeightDecay', 'learning_rate': 1e-05, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False, 'weight_decay_rate': 0.01}
• training_precision: float32

Training results

Train Loss	Train Accuracy	Train Wermet	Validation Loss	Validation Accuracy	Validation Wermet	Epoch
5.4192	0.0107	1.9359	3.9929	0.0112	3.4029	0
4.7175	0.0116	1.3557	3.9525	0.0113	3.2613	1
4.6756	0.0117	1.4198	3.9189	0.0113	2.6795	2
4.6543	0.0117	1.3165	3.9021	0.0114	2.2678	3
4.6317	0.0118	1.2794	3.8796	0.0114	1.8964	4
4.6128	0.0118	1.2033	3.8579	0.0115	1.6353	5
4.5945	0.0118	1.1814	3.8787	0.0114	3.6041	6
4.5719	0.0119	1.1171	3.8418	0.0116	1.1922	7
4.5503	0.0120	1.1435	3.8061	0.0117	1.8502	8
4.5235	0.0121	1.0483	3.7736	0.0118	1.4279	9

Framework versions

• Transformers 4.33.0.dev0
• TensorFlow 2.13.0
• Tokenizers 0.13.3