whisper-medium-swagen-female-model

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

• Loss: 0.5405
• Wer: 0.3398

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: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• 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: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.4291	0.4700	200	0.8217	0.4967
0.9176	0.9401	400	0.6348	0.4266
0.5492	1.4089	600	0.5868	0.4094
0.6243	1.8790	800	0.5535	0.3275
0.2196	2.3478	1000	0.5643	0.3577
0.2211	2.8179	1200	0.5405	0.3398
0.0999	3.2867	1400	0.5826	0.3283
0.1111	3.7568	1600	0.5537	0.3277
0.0423	4.2256	1800	0.6012	0.3188

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0