whisper-base-300v2

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

• Loss: 1.0748
• Wer Ortho: 86.4865
• Wer: 86.4865

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: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant_with_warmup
• lr_scheduler_warmup_steps: 30
• training_steps: 300

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Ortho	Wer
0.129	20.0	60	1.0256	72.9730	72.9730
0.0005	40.0	120	1.0519	81.0811	81.0811
0.0001	60.0	180	1.0650	86.4865	86.4865
0.0001	80.0	240	1.0748	86.4865	86.4865

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1