whisper-base-finetuned-200

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: nan
• Wer Ortho: 100.0
• Wer: 108.1081

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: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant_with_warmup
• lr_scheduler_warmup_steps: 20
• training_steps: 200
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Ortho	Wer
0.0	13.3333	40	nan	100.0	108.1081
0.0	26.6667	80	nan	100.0	108.1081
0.0	40.0	120	nan	100.0	108.1081
0.0	53.3333	160	nan	100.0	108.1081
0.0	66.6667	200	nan	100.0	108.1081

Framework versions

• Transformers 4.41.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1