Dev372/Medical_tiny_en_1_0v_check_train

English Whisper Model

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

• Loss: 0.1019
• Wer: 6.0845

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: 18
• 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: 1500
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.8606	1.0101	100	1.6059	9.1965
1.0954	2.0202	200	1.0513	6.2935
0.6627	3.0303	300	0.6161	6.6187
0.1747	4.0404	400	0.1527	4.2034
0.0554	5.0505	500	0.0924	5.3182
0.028	6.0606	600	0.0819	4.0641
0.0132	7.0707	700	0.0835	5.6897
0.0085	8.0808	800	0.0913	5.5504
0.0056	9.0909	900	0.0932	5.8059
0.0043	10.1010	1000	0.0982	5.8059
0.0023	11.1111	1100	0.0990	5.8755
0.0016	12.1212	1200	0.1005	6.0149
0.0015	13.1313	1300	0.1015	6.0613
0.0016	14.1414	1400	0.1017	6.0613
0.0015	15.1515	1500	0.1019	6.0845

Framework versions

• Transformers 4.42.4
• Pytorch 2.1.2
• Datasets 2.20.0
• Tokenizers 0.19.1