Here are some cases for learning.
Each case can be run independently.
This demo shows a basic LSTM+CRF structure that is based on the code provided by pytorch tutorial file.
This demo is a varant of LSTM_CRF_pytorch_demo in which the task it's facing is no longer NER but CET(College English Test).
This demo uses the python package difflib to deal with the data.
This demo uses a CNN to train the minst dataset, and the model can predict a number written by human.