nocturneFlow/whisper-small-fleurs-augmented-kk

Whisper Small KK - Kazakh - Fleurs Augmented

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

• Loss: 0.2814
• Wer: 22.2891

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 OptimizerNames.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: 200
• training_steps: 1000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.3714	1.0	200	0.2556	23.7164
0.2797	2.0	400	0.2565	23.1212
0.1609	3.0	600	0.2644	22.2688
0.084	4.0	800	0.2738	22.3906
0.059	5.0	1000	0.2814	22.2891

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu118
• Datasets 3.3.2
• Tokenizers 0.21.0