How to train tensorflow chat application on updated dataset everyday? Q: I have created chatbot on Cornell movie dataset and it's working fine. I have trained chatbot application up to global step 330000. I am using tensorflow library. I am also receiving output fine. Then I created this new dataset file by modifying original dataset and i wish to train chatbot application with updated files. Now should I delete previously saved checkpoints and saved data and start training from zero or should i train from 330000 onwards without worrying changes in dataset.