xmzhu/whisper-tiny-zh-TW

Whisper Tiny Chinese (Taiwanese Mandarin)

This model is a fine-tuned version of xmzhu/whisper-tiny-zh on the mozilla-foundation/common_voice_11_0 zh-TW dataset. It achieves the following results on the evaluation set:

• Loss: 0.4879
• Wer: 68.8434

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: 64
• eval_batch_size: 32
• 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	Wer
0.1221	6.02	1000	0.4879	68.8434
0.0119	13.02	2000	0.5567	70.2354
0.004	20.01	3000	0.5890	70.6244
0.0027	27.0	4000	0.6128	72.4053
0.0021	33.02	5000	0.6177	71.9140

Framework versions

• Transformers 4.26.0.dev0
• Pytorch 1.13.1+cu117
• Datasets 2.7.1.dev0
• Tokenizers 0.13.2