polixonrio/whisper-small-fy-NL-Transfer-From-English

Whisper Small Western Frisian (Netherlands)

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

• Loss: 0.5703
• Wer: 21.8466

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
0.0078	10.01	1000	0.5184	23.0973
0.0009	21.0	2000	0.5653	22.5434
0.0007	31.01	3000	0.5703	21.8466
0.0004	42.0	4000	0.5968	21.9574
0.0003	52.01	5000	0.6044	22.0360

Framework versions

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