SmartWhatt/rangsitra-whisper-sm-1-1-10k-rgss

Rangsitra Whisper

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

• Cer: 11.8774
• Wer: 39.2174
• Loss: 0.3022

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: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: warmup_stable_decay
• lr_scheduler_warmup_steps: 312
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Cer	Wer	Validation Loss
0.4393	1.0	625	14.2863	47.6737	0.3219
0.2772	2.0	1250	12.7028	43.0088	0.2889
0.1257	3.0	1875	12.7212	42.2397	0.2888
0.0442	4.0	2500	12.0394	40.2961	0.2970
0.0166	5.0	3125	11.8774	39.2174	0.3022

Framework versions

• Transformers 4.52.4
• Pytorch 2.7.1+cu126
• Datasets 3.6.0
• Tokenizers 0.21.2