Hi @Suteja Kanuri​ ,

Thank you for the solution. I have also noticed that instead of passing pickle files path, I can pass the fitted_tokenizer and fitted_label_encoder object directly to the KerasModel class. This solution worked for me. But yours also looks correct!