gwanju_large_model

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

• Loss: 0.5238
• Cer: 238.8739

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: 20000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Cer
0.2792	1.4819	5000	0.3803	453.3834
0.1361	2.9638	10000	0.3823	336.9865
0.0335	4.4458	15000	0.4644	344.2898
0.0097	5.9277	20000	0.5238	238.8739

Framework versions

• Transformers 4.41.2
• Pytorch 2.2.2+cu121
• Datasets 2.19.2
• Tokenizers 0.19.1