Week2 Assignment 2 - Emojify

Course Q&A Deep Learning Specialization Sequence Models
1

I have some confuse point related to this assignment:
1- If in the input sentence have words, or specifics word is not in glow.6B.50d.txt
That mean we can not word to index of that word.
What will happen?
2- How to save this model for next time use, no need train again?
3- I Have a database. I want to make a chat bot to retrieve the information form that database by some limited category (as Y label). Because is technical information, It have many slang words/ special words that is not in glow.6B.50d.txt. Could I use this method?

Please support me!

2
  1. There are a number of points in the lectures in Week 2 where Prof Ng discusses how to deal with words that are not in the vocabulary. You don’t have an embedding for them, so the typical method is to just drop or ignore them. I suggest you review the lecture notes for Week 2 for more info.
  2. Saving and loading models is a standard thing provided by TF/Keras APIs. Try googling “Save Keras Model” or “Load Keras Model”. Googling “save keras model” gets me this TF tutorial as the top level hit. Are you feeling lucky today? :nerd_face:
  3. Well, maybe you need to consider using a word embedding database that is a better match for your problem. There are quite a few pretrained models available. Or maybe you can try that model and with the handling of non-vocabulary words (e.g. just ignoring them) it will still work well enough.
3

I understood. Thanks for always support me.:blush: