zuazo/whisper-small-gl-train

Whisper Small Galician

This model is a fine-tuned version of openai/whisper-small on the mozilla-foundation/common_voice_13_0 gl dataset. It achieves the following results on the evaluation set:

• Loss: 0.2772
• Wer: 10.2463

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
0.2733	5.81	1000	0.2584	12.1814
0.1012	11.63	2000	0.2476	10.6250
0.0514	17.44	3000	0.2628	10.3500
0.0299	23.26	4000	0.2739	10.3189
0.0235	29.07	5000	0.2772	10.2463

Framework versions

• Transformers 4.37.2
• Pytorch 2.2.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.1