zuazo/whisper-medium-eu-train

Whisper Medium Basque

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

• Loss: 0.3985
• Wer: 14.1127

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: 10000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.1127	5.85	1000	0.2776	17.5623
0.0225	11.7	2000	0.3129	15.6320
0.0074	17.54	3000	0.3277	14.9530
0.0041	23.39	4000	0.3551	14.8018
0.0032	29.24	5000	0.3698	14.6245
0.0019	35.09	6000	0.3877	14.6084
0.0014	40.94	7000	0.3891	14.4976
0.0008	46.78	8000	0.3946	14.2759
0.0007	52.63	9000	0.3987	14.3182
0.0005	58.48	10000	0.3985	14.1127

Framework versions

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