Goal : 在特定目的下，為文件做出小改動，快速第1次謄寫

實做CBOW文件，tensorflow_word2vec_cbow_basic.py

(柏翰)Step 1: Download the data.

(北極)Step 2: Build the dictionary and replace rare words with UNK token.

(立先)Step 3: Function to generate a training batch for the skip-gram model.

(家鴻)Step 4: Build and train a skip-gram model.

(家鋒)Step 5: Begin training.

Step 6: Visualize the embeddings.

[0]https://gist.github.com/yxtay/a94d971955d901c4690129580a4eafb9

***在寫CBOW有用到 tf. 的函式，就可以去查找
https://www.tensorflow.org/api_docs/python/tf

10/21 家鋒 & 北極 介紹numpy

10/28 團隊介紹CBOW程式