jeff7522553/whisper-small-gy

Whisper small TW - from gy

This model is a fine-tuned version of openai/whisper-medium on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1443
• Cer: 215.2740

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: 16
• 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: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Cer
0.1167	1.02	1000	0.1762	98.2782
0.0271	2.04	2000	0.1489	127.0736
0.0148	3.06	3000	0.1434	144.6924
0.0059	4.08	4000	0.1438	162.0580
0.0018	5.1	5000	0.1443	215.2740

Framework versions

• Transformers 4.37.2
• Pytorch 2.5.1+cu121
• Datasets 3.5.0
• Tokenizers 0.15.2