mrmuminov/whisper-large-v3-turbo

Whisper Large v3 Turbo - Bahriddin Muminov

This model is a fine-tuned version of openai/whisper-large-v3-turbo on the Common Voice 16.1 dataset. It achieves the following results on the evaluation set:

• Loss: 0.2643
• Wer: 25.0740

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1000
• training_steps: 10000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.3109	0.04	2000	0.4220	36.8942
0.2529	0.07	4000	0.3593	31.0915
0.2123	0.11	6000	0.3150	28.2694
0.1936	0.14	8000	0.2773	27.4353
0.1716	0.18	10000	0.2643	25.0740

Framework versions

• Transformers 4.37.2
• Pytorch 2.2.2+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2