Steps

[esthicodes]

1. Dataset preparation - put your data into a csv file or json file
2. Tokensize / word embedding, etc.
3. NLP model training - RNN?
4. Build a server and load the model there - using Python libraries like Flask
5. Make APIs, so one can send texts to the server (e.g. Swiss German sentences) and get the translation back (e.g. Chinese)
6. Build an app to show an interface, e.g. one can choose which language to translate from and to, enter sentences, and a button to send the information to the server, receive the translation, and show it in UI