jlondonobo/whisper-small-pt

Whisper Small Portuguese 🇧🇷🇵🇹

This model is a fine-tuned version of openai/whisper-small on the mozilla-foundation/common_voice_11_0 pt dataset. It achieves the following results on the evaluation set:

• Loss: 0.3887
• Wer: 10.2453

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: 64
• eval_batch_size: 32
• 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: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0492	3.04	1000	0.2459	10.3562
0.0065	7.02	2000	0.3180	10.4521
0.002	11.0	3000	0.3571	10.2924
0.0009	14.04	4000	0.3816	10.2268
0.0008	18.02	5000	0.3887	10.2453

Framework versions

• Transformers 4.26.0.dev0
• Pytorch 1.13.0+cu117
• Datasets 2.7.1.dev0
• Tokenizers 0.13.2