MR-Eder/GRAG-whisper-large-v3-turbo-german-v1

Whisper Small German - GRAG

This model is a fine-tuned version of openai/whisper-large-v3-turbo on the GER-TTS-50-Conversations dataset. It achieves the following results on the evaluation set:

• Loss: 0.4391
• Wer: 15.1703

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 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: 500
• training_steps: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.066	8.3333	1000	0.3653	15.4640
0.0038	16.6667	2000	0.4180	15.0235
0.0006	25.0	3000	0.4340	15.1882
0.0004	33.3333	4000	0.4391	15.1703

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.0+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3