whisper_input_decoder_shift_r_labels_no_force__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: 3.6757
• Train Accuracy: 0.0136
• Train Wermet: 0.7548
• Validation Loss: 3.3141
• Validation Accuracy: 0.0116
• Validation Wermet: 0.8400
• 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
5.6348	0.0091	1.5865	4.2935	0.0093	0.9579	0
4.9212	0.0099	0.9054	4.1262	0.0097	0.9390	1
4.6819	0.0107	0.8319	3.9071	0.0103	0.8966	2
4.4443	0.0114	0.8310	3.7367	0.0106	0.8939	3
4.2479	0.0119	0.8226	3.6101	0.0109	0.8696	4
4.0911	0.0124	0.8103	3.5364	0.0110	0.8946	5
3.9590	0.0127	0.7913	3.4556	0.0113	0.8388	6
3.8513	0.0130	0.7794	3.4106	0.0114	0.8515	7
3.7607	0.0133	0.7657	3.3507	0.0115	0.8261	8
3.6757	0.0136	0.7548	3.3141	0.0116	0.8400	9

Framework versions

• Transformers 4.34.0.dev0
• TensorFlow 2.13.0
• Tokenizers 0.13.3