zuazo/whisper-base-eu-train

Whisper Base Basque

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

• Loss: 0.5071
• Wer: 26.5933

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: 2.5e-05
• train_batch_size: 128
• eval_batch_size: 64
• 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
0.3263	11.63	1000	0.5019	33.1701
0.1194	23.26	2000	0.4647	28.2859
0.0576	34.88	3000	0.4863	27.2401
0.0337	46.51	4000	0.5028	26.6135
0.0276	58.14	5000	0.5071	26.5933

Framework versions

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