jwh1449/whisper_base_KoSpeech

whisper_small_KoSpeech

This model is a fine-tuned version of openai/whisper-base on the AIhub_KoSpeech_dataset2 dataset. It achieves the following results on the evaluation set:

• Loss: 0.4259
• Cer: 13.9789

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: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• 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	Cer
0.464	0.1580	1000	0.4912	16.0865
0.4841	0.3161	2000	0.4548	14.1104
0.4447	0.4741	3000	0.4352	13.7069
0.4254	0.6321	4000	0.4259	13.9789

Framework versions

• Transformers 4.52.0.dev0
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1