###### tags: `人工智慧` # 3.TensorFlow + Keras 的第一步 新手開始安裝(Windows) === #### 建立 trnsorflow anaconda虛擬環境 ``` conda create --name tensorflowAI python=3.5 anaconda ``` #### 啟動 anaconda虛擬環境 ``` activate tensorflowAI ``` #### 關閉 anaconda虛擬環境 ``` deactivate tensorflowAI ``` --- #### 安裝 tensorflow CPU ``` pip install tensorflow ``` :::info :mega: 出現錯誤 ERROR: Cannot uninstall 'wrapt'. It is a distutils installed project and thus we cannot accurately determine which files belong to it which would lead to only a partial uninstall. ``` pip install -U --ignore-installed wrapt enum34 simplejson netaddr ``` :mega: python版本與anaconda虛擬環境不一樣 ``` conda install tensorflow ``` ::: #### 安裝 keras ``` pip install keras ``` :::info :mega: python版本與anaconda虛擬環境不一樣 ``` conda install keras ``` ::: --- #### 啟動jupyter ``` activate tensorflowAI jupyter notebook ``` #### 讓jupyter更好用 [jupyter themes](https://www.796t.com/article.php?id=18214) [Jupyter extend](https://zhuanlan.zhihu.com/p/35430879) [jupyter 補字](https://www.796t.com/article.php?id=217867) [補](https://github.com/codota/jupyter-tabnine) :::info :mega: pip沒反應可以試試conda ``` conda install -c conda-forge jupyter_nbextensions_configurator conda install -c conda-forge jupyter_contrib_nbextensions jupyter nbextensions_configurator enable --user jupyter contrib nbextension install --user jt -t oceans16 -f fira -fs 18 -cellw 99% -tfs 20 -dfs 16 -ofs 16 -lineh 170 -T -N -kl ``` ::: [](https://chrislee0728.medium.com/%E4%BA%94%E5%80%8B%E5%A5%BD%E7%94%A8%E7%9A%84-jupyter-notebook-%E6%8F%92%E4%BB%B6-e391b82638d5) --- 新手開始安裝(Linux) === #### [anaconda下載](https://www.anaconda.com/products/individual#Downloads) ``` wget https://repo.anaconda.com/archive/Anaconda3-2021.05-Linux-x86_64.sh ``` #### 安裝anaconda ``` bash Anaconda3-2021.05-Linux-x86_64.sh -b sudo gedit ~/ .bashrc export PATH="/home/user/anaconda3/bin:$PATH" source ~/.bashrc python --version ``` :notebook: -b batch 是批次安裝 #### anaconda加到PATH ``` sudo gedit ~/ .bashrc export PATH="/home/user/anaconda3/bin:$PATH" ``` #### 重新登入 ``` source ~/.bashrc ``` #### 確定版本 ``` python --version ``` --- #### 安裝 tensorflow CPU ``` pip install tensorflow ``` #### 安裝 keras ``` pip install keras ``` --- #### 啟動jupyter ``` mkdir -p ~/pywork cd ~/pywork jupyter notebook ``` --- # 更方便使用 anacond 跟 jupyter #### Step 1. 安裝[Windows terminal](https://www.microsoft.com/store/productId/9N0DX20HK701) #### Step 2. 搜尋anacond > 右鍵開啟檔案位置 ![](https://i.imgur.com/Gkz4C7Y.png) #### Step 3. 在anacond右鍵開起屬性 > 複製Target ![](https://i.imgur.com/8zhXtU7.png) #### Step 4. 開啟Windows terminal 設定 ![](https://i.imgur.com/7ljKTAS.png) #### Step 5. 新增一個新的 ![](https://i.imgur.com/DAuTEQH.png) --- #### Step 6. 修改設定檔 + 名稱可以任意取 + 將複製的Target貼到命令列 ``` %windir%\System32\cmd.exe "/K" C:\Users\55614\anaconda3\Scripts\activate.bat C:\Users\55614\anaconda3\envs\tensorflow35 ``` + 刪除/修改Target一些內容 ``` cmd.exe /K C:\Users\55614\anaconda3\Scripts\activate.bat C:\Users\55614\anaconda3\envs\tensorflow35 ``` + 圖示 ``` C:\Users\55614\anaconda3\Menu\anaconda-navigator.ico ``` [guid]::NewGuid() + 如圖 ![](https://i.imgur.com/fHhcAh2.png) #### 新增jupyter 也一樣重複 Step 2. ![](https://i.imgur.com/wt1Ijgy.png) --- #### 完成就可以直接呼叫了 ![](https://i.imgur.com/Y8cw4Dt.png)