Hiperparameters

#9
by jigmam - opened

Hello, when I try to extract text from some images, the model repeats sentences. For example: '''CATASTAL DEL MUNICIPAL

REPERTORIO CATASTAL DEL MUNICIPAL

REPERTORIO CATASTAL DEL MUNICIPAL

REPERTORIO CATASTAL DEL MUNICIPAL

REPERTORIO CATASTAL DEL MUNICIPAL

REPERTORIO CATASTAL DEL MUNICIPAL

REPERTORIO CATASTAL DEL MUNICIPAL

''' I know my text is in Spanish, but are there any hyperparameters to improve or avoid the model repeating sentences?

Sign up or log in to comment