This model is a fine-tuned version of openai/whisper-base on the mangoo111/2025re_learn dataset.
It achieves the following results on the evaluation set:
Loss: 1.5861
Cer: 26.9608
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 adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
