Cleaner Names Model

This model was trained using information from the google places API (10K+). These restaurants are from countries like Mexico and Colombia, so spanish is principal language. The idea is remove automatically the stopwords in the name.

The model was trained using an specific prompt, you will have to adapt the name in this way:

""" REGULAR NAME: {name}. CLEANED NAME: """ For better results use the name in upper case.

Examples:

• POB MCDONALDS POBLADO HAMBURGUESA
• BURGER KING SHOPPING RECIFE 15883
• CENTRO COMERCIAL GRAN ESTACION, DOMINOS PIZZA

Using the prompt: REGULAR NAME: CENTRO COMERCIAL GRAN ESTACION, DOMINOS PIZZA. CLEANED NAME:

On the Eval Dataset the Bleu Score is higher than 0.7. Its good, but can be better.