didiudom94/whisper-med-kor-en

Whisper Small Ko to En

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

• Loss: 1.4544
• Bleu: 0.1211

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: 8
• 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	Bleu
1.6647	0.2253	1000	1.6390	0.0927
1.5514	0.4507	2000	1.5395	0.1087
1.5369	0.6760	3000	1.4836	0.1153
1.4714	0.9013	4000	1.4544	0.1211

Framework versions

• Transformers 4.45.2
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.20.3