zagibest/whisper-medium-custom-data

Whisper Medium MN with custom data - Zagi

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

• Loss: 0.0918
• Wer: 10.8352

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: 1
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 4000

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.5144	0.15	500	0.3790	43.5855
0.3922	0.3	1000	0.2215	26.4686
0.2435	0.46	1500	0.1774	21.2074
0.2275	0.61	2000	0.1451	18.1786
0.1447	0.76	2500	0.1279	15.7240
0.2028	0.91	3000	0.1065	13.0327
0.1068	1.06	3500	0.1002	12.2796
0.087	1.21	4000	0.0918	10.8352

Framework versions

• Transformers 4.39.1
• Pytorch 2.0.1+cu117
• Datasets 2.18.0
• Tokenizers 0.15.2