# IIIEDU-BDSE20 Introduction to Deep Learning

記得以下網址, 方便之後回來此頁面: https://hackmd.io/@chweng/bdse-20

---
## 資料下載
* 2021/08/24
* [[講義, 資料集, 部分範例程式]](https://www.dropbox.com/t/7vOHOtgiGRJBIJzJ) (下載密碼: 20bdse)
* 2021/09/01
* [[三個範例筆記本]](https://www.dropbox.com/t/KRTAJnrROaCgDP5r) (下載密碼: 20bdse02)
* 2021/09/04
* [[Jupyter Notebooks (html)]](https://drive.google.com/drive/folders/1X3z9bt29a0bYEoay-ReWDHX4SKq0pw0w?usp=sharing)
* 2021/09/06
* [[一個範例筆記本]](https://www.dropbox.com/t/SUw2MiJMc9ON3Cz3)
* 2021/09/14
* [[兩本電子書]](https://www.dropbox.com/sh/9qqomso5mogvqhl/AADb1yLkiz_tysPskQz5Ww5Sa?dl=1)
## 補充投影片
* 2021/08/28
* 2021/09/01
* 2021/09/04

## 補充資訊
* [PyTorch's Cross Entropy Loss](https://pytorch.org/docs/stable/generated/torch.nn.CrossEntropyLoss.html)
* 多GPU/CPU訓練 (distributed training)
* [Uber Horovod](https://github.com/horovod/horovod)
* [[Spark + TensorFlow]?](http://blog.madhukaraphatak.com/tensorflow-on-spark-3.0/)
* Python Machine Learning 課本的[筆記本教材](https://github.com/rasbt/python-machine-learning-book-2nd-edition)
* Stanford CS231 [[Convolution layer]](https://cs231n.github.io/convolutional-networks/)
* The BatchNorm Layer [[TF Keras API]](https://keras.io/api/layers/normalization_layers/batch_normalization/) [[PyTorch API]](https://pytorch.org/docs/stable/generated/torch.nn.BatchNorm2d.html)

## 環境設定
1. 設定NVSMI (看NVIDIA DRIVER有沒有裝好)
1. 將```C:\Program Files\NVIDIA Corporation\NVSMI```加入System ```PATH```
2. 加入後, 開啟終端機(如GIT BASH), 然後輸入```nvidia-smi.exe```, 按Enter, 即可顯示出GPU使用率等資訊
2. 安裝CUDA
3. 安裝cuDNN
將解壓縮後的資料, 逐一複製到```C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.1```
4. 確定nvcc有安裝 (看CUDA有沒有裝好)

5. 安裝PyTorch
   https://pytorch.org/get-started/locally/
   安裝好後, 執行python, 然後載入PyTorch並且嘗試看能否印出PyTorch版本:

import torch
print(torch.__version__)

6. 安裝TensorFlow: https://www.tensorflow.org/install/gpu?hl=zh-tw

import tensorflow as tf
print(tf.__version__)

---
請安裝Anaconda後，加裝如下套件:
* 基本套件

pip install pandas matplotlib seaborn numpy scipy

* 機器學習套件  

pip install sklearn

* 其它套件

pip install pillow graphviz
pip install mlxtend
pip install numexpr

conda install -y shapely
pip install imgaug opencv-python==4.4.0.46

pip install jupyterlab

---
翁啟閎
[e-mail]