Whisper Turbo Ko v0.0.2

This model is a fine-tuned version of openai/whisper-large-v3-turbo on the Common Voice 17.0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.5527
• Cer: 10.4193

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
• distributed_type: multi-GPU
• num_devices: 2
• total_train_batch_size: 32
• total_eval_batch_size: 16
• 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
0.0024	43.4783	1000	0.5080	11.2617
0.0001	86.9565	2000	0.5336	10.3515
0.0001	130.4348	3000	0.5476	10.3031
0.0	173.9130	4000	0.5527	10.4193

Framework versions

• Transformers 4.52.4
• Pytorch 2.7.1+cu126
• Datasets 3.6.0
• Tokenizers 0.21.1