Jupyter 設定

Mon, 09 Aug 2021 07:48:37 +0800

Install

pip install jupyter

sudo pip3 install --upgrade pip
sudo pip3 install jupyter

Error

ModuleNotFoundError: No module named '_sqlite3'
ModuleNotFoundError: No module named 'pysqlite2'

Used

start

cd /要用來存資料的資料夾/
jupyter notebook

參數

--ip
--port
--no-browser
--help

config

環境設定

ipython --ipython-dir= # override the default IPYTHONDIR directory, ~/.ipython/ by default
ipython profile create foo # create the profile foo
ipython profile locate foo # find foo profile directory, IPYTHONDIR by default,
ipython --profile=foo # start IPython using the new profile

產生設定檔

jupyter –-generate-config

會在產生 ~/.jupyter/jupyter_notebook_config.py 設定檔

jupyter server

產生設定檔

jupyter notebook --generate-config

產生密碼

jupyter notebook password

會在 jupyter_notebook_config.json 存密碼的 hash value，需要把它丟到 jupyter_notebook_config.py 裡的 c.NotebookApp.password =

修改設定檔

# Set options for certfile, ip, password, and toggle off

# Set ip to '*' to bind on all interfaces (ips) for the public server
c.NotebookApp.ip = '*'
c.NotebookApp.password = u'sha1:bcd259ccf...<your hashed password here>'
c.NotebookApp.open_browser = False

# It is a good idea to set a known, fixed port for server access
c.NotebookApp.port = 9999

Jupyter Daemon

檔案：/lib/systemd/system/jupyter.service

[Unit]
Description=Jupyter Notebook

[Service]
Type=simple
PIDFile=/run/jupyter.pid
# Step 1 and Ste 2 details are here..
# -----------------------------------
ExecStart=/usr/local/bin/jupyter-notebook --config=/home/calee/.jupyter/jupyter_notebook_config.py
User=calee
Group=calee
WorkingDirectory=/home/calee/Project/Jupyter
Restart=always
RestractSec=10
#KillMode=mixed

[Install]
WantedBy=multi-user.target

執行

sudo systemctl start jupyter.service
sudo systemctl daemon-reload
sudo systemctl restart jupyter.service
sudo systemctl enable jupyter.service

NGINX proxy setting

位置：/etc/nginx/sites-enabled/notebook.conf

upstream notebook {
 server localhost:8888;
}
server{
 listen 80;
 server_name notebook.calee.xyz;


 location / {
 proxy_pass http://notebook;
 proxy_set_header Host $host;
 }

 location ~ /api/kernels/ {
 proxy_pass http://notebook;
 proxy_set_header Host $host;
 # websocket support
 proxy_http_version 1.1;
 proxy_set_header Upgrade "websocket";
 proxy_set_header Connection "Upgrade";
 proxy_read_timeout 86400;
 }
 location ~ /terminals/ {
 proxy_pass http://notebook;
 proxy_set_header Host $host;
 # websocket support
 proxy_http_version 1.1;
 proxy_set_header Upgrade "websocket";
 proxy_set_header Connection "Upgrade";
 proxy_read_timeout 86400;
 }
}

Restart Nginx

sudo systemctl restart nginx

TensorFlow 亂記

Mon, 09 Aug 2021 07:48:37 +0800

安裝

CPU

使用 pip 安裝

sudo apt-get install python3 python3-pip python3-dev
pip3 install tensorflow
sudo pip3 install --upgrade

docker

sudo apt-get install docker
sudo groupadd docker
sudo usermod -aG docker $USER

docker run -it -p [hostPort]:[containerPort] [TensorFlowCPUImage]
docker run -it -p 8888:8888 gcr.io/tensorflow/tensorflow

GPU support

不一定要裝，但是可以加速
pip

sudo apt-get install python3-pip python3-dev
pip3 install tensorflow-gpu
sudo pip3 install --upgrade

docker

nvidia-docker run -it -p [hostPort]:[containerPort] [TensorFlowGPUImage]
nvidia-docker run -it -p 8888:8888 gcr.io/tensorflow/tensorflow:latest-gpu

computational graph

computational graph 是由許多 TensorFlow 運算節點（nodes）所組成的運算藍圖，每個節點可以接受任意個數的 tensors 作為輸入資料（或是沒有任何輸入也可以），並輸出一個 tensor

建立兩個節點
```
node1 = tf.constant(3.0, dtype=tf.float32)
node2 = tf.constant(4.0) # also tf.float32 implicitly
print(node1, node2)
```
可以用 dtype 指定型別，預設型別是 float32 印出來的並不會是 3.0 跟 4.0，因為他們事實上並不是 value，而是 node。只有在 evaluated 時，才會呈現出 3.0 跟 4.0，

API

Tensors

基礎的資料類型(class/data type)，它是一種多維度的陣列，其陣列的維度稱為 rank

Placeholder

placeholder 是一種可以讓 computational graph 保留輸入欄位的節點，其允許實際的輸入值留到後來再指定

Variable

參數的部份我們可以透過 variable 的節點來指定

Python | Chia-An Lee