wwwtwwwt/whisper-tiny-NewsPolitics-Social-Commentary

Whisper Tiny En - NewsPolitics - Social Commentary

This model is a fine-tuned version of openai/whisper-tiny on the fineaudio-NewsPolitics-Social Commentary dataset. It achieves the following results on the evaluation set:

• Loss: 0.9274
• Wer: 52.6551

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: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.8927	0.5705	1000	1.0172	59.8874
0.702	1.1409	2000	0.9510	56.5795
0.6797	1.7114	3000	0.9338	55.1387
0.6021	2.2818	4000	0.9274	52.6551

Framework versions

• Transformers 4.45.2
• Pytorch 2.4.0
• Datasets 3.1.0
• Tokenizers 0.20.0