KrorngAI/whisper-small-finetuned-km

Whisper Small Kh - Kimang Khun

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

• Loss: 0.1933
• Wer: 59.5835

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.SGD and the args are: No additional optimizer arguments
• lr_scheduler_type: constant
• lr_scheduler_warmup_steps: 15
• num_epochs: 2
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0863	0.5587	100	0.2045	62.8281
0.0796	1.1173	200	0.2013	60.6320
0.0798	1.6760	300	0.1996	60.7927

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu126
• Datasets 4.2.0
• Tokenizers 0.22.1