/stanford-tensorflow-tutorials

Deprecated.

Primary LanguagePython

tf-stanford-tutorials

This repository contains code examples for the course CS 20SI: TensorFlow for Deep Learning Research.
It will be updated as the class progresses.
Detailed syllabus and lecture notes can be found here http://cs20si.stanford.edu

这里是斯坦福大学2017年课程，CS 20SI《深度学习研究中的 TensorFlow》所有示例和作业的代码。
课程大纲、课堂笔记和PPT可以在 https://web.stanford.edu/class/cs20si/syllabus.html 找到。

Quickstart

  1. Install Python and Jupyter notebooks
  2. Install TensorFlow
  3. Clone the repository
  4. Find and open examples folder with Jupyter notebook
  5. Start with jupyter_02_feed_dict.ipynb

开始学习

  1. 安装 Python 和 Jupyter notebooks
  2. 安装 TensorFlow
  3. 克隆这个项目
  4. 在 Jupyter notebook 中找到 examples 文件夹
  5. 打开 jupyter_02_feed_dict.ipynb 开始学习

Jupyter Notebooks

We are adding Jupyter notebooks for examples and assignments, with names "jupyter_blah_blah.ipynb" under /assignments and /examples. You can now open up the notebook and execute and debug your TensorFlow code line by line! The converted notebooks are labelled with checks below.
We also added the inline TensorBoard code, a useful tool for visualizing and debugging your graph. See show_tf_graph.py for instructions. The conversion is still in progress, and we welcome any pull requests!

为了方便学习，我们正在把代码改写为 Jupyter notebooks ，它们位于 assignments 和 examples 文件夹下。通过 Jupyter notebooks，我们可以一行行地实现和调试 TensorFlow 的代码。已经改写的代码会在下面的代码列表中标识。
此外，Jupyter notebook 还支持实时查看 TensorBoard，详情参考 show_tf_graph.py
项目正在进行，欢迎提交pull request！

Install TensorFlow, 安装 TensorFlow

Windows users often have trouble installing TensorFlow with GPU support on Windows. I have found a useful step-by-step installation. Also, take a look at @mrry's self check code if you have trouble.
Windows 用户在安装支持GPU的TensorFlow过程中经常遇到问题，下面两个链接分别给出了逐步的安装教程和系统自检代码：
https://nitishmutha.github.io/tensorflow/2017/01/22/TensorFlow-with-gpu-for-windows.html https://gist.github.com/mrry/ee5dbcfdd045fa48a27d56664411d41c

Models include:

Code with checks indicate Jupyter notebooks.

In the folder "examples":

  • Linear Regression with Chicago's Fire-Theft dataset, 线性回归
  • Logistic Regression with MNIST, 逻辑回归
  • Word2vec skip-gram model with NCE loss, word2vec语言模型
  • Convnets with MNIST, 卷积网络
  • Autoencoder (by Nishith Khandwala), 自编码机
  • Deepdream (by Jon Shlens), Deepdream
  • Character-level language modeling, 字符级生成语言模型

In the folder "assignments":

  • Style Transfer, 风格迁移
  • Chatbot using sequence to sequence with attention, 基于seq2seq与注意力机制的聊天机器人

Misc

  • Examples on how to use data readers, TFRecord
  • Embedding visualization with TensorBoard
  • Usage of summary ops
  • Exercises to be familiar with other special TensorFlow ops
  • Demonstration of the danger of lazy loading
  • Convolutional GRU (CRGU) (by Lukasz Kaiser)

Note (as of July 11, 2017)

I've updated the code to TensorFlow 1.2 and Python3, except the code for chatbot. I will update the code for chatbot soon.