whisper-large-v3-polish-impaired

This model is a fine-tuned version of openai/whisper-large-v3 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.5641
• Wer: 83.0040

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: 50
• training_steps: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
2.239	0.05	1	2.7489	84.5850
1.8286	0.1	2	2.7304	84.5850
2.1458	0.15	3	2.6956	84.5850
2.5037	0.2	4	2.6399	83.3992
2.5533	0.25	5	2.5641	83.0040

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cpu
• Datasets 3.2.0
• Tokenizers 0.21.0