whisper-medium-bigcgen-male-5hrs-model

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

• Loss: 0.6817
• Wer: 0.5054

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: 1.75e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 4
• 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
3.6678	0.6211	200	0.8233	0.6018
2.3045	1.2422	400	0.7233	0.6420
2.7509	1.8634	600	0.6881	0.5239
1.7043	2.4845	800	0.6817	0.5054
0.8068	3.1056	1000	0.7271	0.4915
0.9957	3.7267	1200	0.7331	0.4893
0.3976	4.3478	1400	0.7714	0.4943

Framework versions

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