harpreet1/whisper-medium-english

Whisper Small en-Harpreet Singh

This model is a fine-tuned version of openai/whisper-medium on the med-audio-3 dataset. It achieves the following results on the evaluation set:

• Loss: 0.0384
• Wer: 12.5604

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: 1
• eval_batch_size: 2
• 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: 100
• training_steps: 500

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0221	0.3268	100	0.0383	13.1643
0.0339	0.6536	200	0.0373	13.0435
0.0265	0.9804	300	0.0382	12.9227
0.0048	1.3072	400	0.0388	13.0435
0.0088	1.6340	500	0.0384	12.5604

Framework versions

• Transformers 4.46.3
• Pytorch 2.5.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3