AC-5

by Incomplete

This is a collection of heavily documented Python files/Jupyter Notebooks containing TensorFlow codes/models.

Each .py file is a standalone module, and comes with a solarized-light-themed .html file for online reading.

The .py file is supposed to be executed interactively, hence is organized using "cells", which is supported by various tools, such as Spyder, PyCharm and Neovim with senter.

All the codes can be found at github.com/qwfy/AC-5.

TOC

• seq2seq_addition.py: Addition with a basic seq2seq model.
  Given a string input "12+15", the model should produce a string output "27".
• attention_word_to_number.ipynb: Tutorial on attention-based seq2seq model.
  It includes both the math and the code. A model is built to translate string like "one hundred and twenty seven" to string "127".
• sneaky_python.ipynb: Some Python gotchas for functional programmers.