adrianSauer/whisper-base-wer-extension

Common Voice 16 - Guarani

This model is a fine-tuned version of openai/whisper-base on the Common Voice 16 dataset. It achieves the following results on the evaluation set:

• Loss: 0.4522
• Wer: 53.9363

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: 2e-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: 3000
• training_steps: 3000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
2.9797	0.4955	500	1.0523	91.7365
1.1502	0.9911	1000	0.6757	87.2138
0.717	1.4866	1500	0.5718	63.4841
0.5485	1.9822	2000	0.5059	58.3473
0.3345	2.4777	2500	0.4828	56.7281
0.2748	2.9732	3000	0.4522	53.9363

Framework versions

• Transformers 4.44.0
• Pytorch 2.3.1+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1