IbratDO/whisper-small-linkdata

Whisper Small UZ - Link Data

This model is a fine-tuned version of openai/whisper-small on the UzbekVoice 1K Row dataset. It achieves the following results on the evaluation set:

• Loss: 0.8197
• Wer: 68.5714

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: 4
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 50
• training_steps: 200
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.4506	1.0	50	1.1262	79.8413
0.7509	2.0	100	0.8611	70.3175
0.3845	3.0	150	0.8157	68.2540
0.2639	4.0	200	0.8197	68.5714

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu126
• Datasets 4.1.1
• Tokenizers 0.22.1