navin-kumar-j/whisper-base-ta

Whisper Base Ta - Navin Kumar J

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

• Loss: 0.2913
• Wer: 1.5234

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: 8
• 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: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.2192	0.2773	1000	0.3592	1.1204
0.2076	0.5546	2000	0.3164	1.1192
0.1881	0.8319	3000	0.2993	1.5272
0.1504	1.1093	4000	0.2913	1.5234

Framework versions

• Transformers 4.51.3
• Pytorch 2.7.0+cu126
• Datasets 3.5.0
• Tokenizers 0.21.1