Krish03/whisper-large-de

Whisper Large De - Krish Kalra

This model is a fine-tuned version of openai/whisper-large-v3-turbo on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.2176
• Wer: 9.2911

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: 5
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.1148	1.0	300	0.1876	10.6676
0.1281	2.0	600	0.2023	11.0805
0.032	3.0	900	0.2043	10.1170
0.0015	4.0	1200	0.2194	9.7729
0.0005	5.0	1500	0.2176	9.2911

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.1+cu124
• Datasets 3.1.0
• Tokenizers 0.20.3