whisper-small-vietnamese

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

• Loss: 0.4479
• Wer: 16.0301
• Cer: 8.5605

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: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant_with_warmup
• lr_scheduler_warmup_steps: 500
• training_steps: 4000

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
0.1455	2.87	1000	0.3513	16.1208	8.5167
0.0198	5.73	2000	0.4002	16.7085	9.2051
0.0059	8.6	3000	0.4233	15.5459	8.2189
0.0049	11.46	4000	0.4479	16.0301	8.5605

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.0
• Datasets 2.14.4
• Tokenizers 0.13.3