jordimas/whisper-medium-ca-5000steps

Whisper Medium Catalan

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

• Loss: 0.2373
• Wer: 10.2052

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: 2
• eval_batch_size: 1
• 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
0.3022	0.2	1000	0.3475	16.1613
0.2221	0.4	2000	0.3142	13.8036
0.2635	0.6	3000	0.2802	12.5606
0.2418	0.8	4000	0.2503	10.9386
0.1958	1.0	5000	0.2373	10.2052

Framework versions

• Transformers 4.26.0.dev0
• Pytorch 1.11.0+cu102
• Datasets 2.8.1.dev0
• Tokenizers 0.13.2