ninninz/whisper-large-v3-croatian-overlap-removed-20

whisper-large-v3-croarian_overlap_removed_20

This model is a fine-tuned version of openai/whisper-large-v3 on the audiofolder dataset. It achieves the following results on the evaluation set:

• Loss: 2.1494
• Wer: 66.4162

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: 16
• eval_batch_size: 8
• 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: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0196	18.18	1000	1.8624	74.2693
0.0034	36.36	2000	2.0306	57.5067
0.0028	54.55	3000	2.1057	61.0400
0.0029	72.73	4000	2.1494	66.4162

Framework versions

• Transformers 4.37.1
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.1