PenguinbladeZ/whisper-small-hk

Whisper Small Canontese X v3

This model is a fine-tuned version of openai/whisper-small on the Common Voice 16.0, the Common Voice 16.1 and the Common Voice 17.0 datasets. It achieves the following results on the evaluation set:

• Loss: 0.2650
• Wer: 55.6319

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: 4
• 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: 3000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0845	0.7087	1000	0.2773	61.5120
0.0285	1.4174	2000	0.2697	56.7010
0.0102	2.1262	3000	0.2650	55.6319

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.0
• Tokenizers 0.19.1