simple_learning Current stage Finish the first version of Bert. Some reinforcement learnings Future plans Add testing stage to the part Add loading pretrained model functionality More improvement