> For the complete documentation index, see [llms.txt](https://purelyvivid.gitbook.io/deep-learning/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://purelyvivid.gitbook.io/deep-learning/conda_env_for_deeplearning_python.md).

# 如何建立一個python深度學習的conda環境

## 安裝 Anaconda

首先你要先裝Anaconda，到[這裡](https://docs.anaconda.com/anaconda/install/)來下載你環境的版本，我的是ubuntu16.04，以下說明都是以這個作業系統環境。

## 建立一個機器學習的基底環境

```
conda create --name ml_base
conda activate ml_base
conda install pandas numpy scipy
conda install scikit-learn
conda install -c conda-forge matplotlib 
conda install -c anaconda git
conda install -c anaconda ipykernel
conda install -c conda-forge jupyterlab
conda deactivate
```

## 建立一個python Tensorflow的conda環境

上述ml\_base環境必須先建製完成

```
conda create --name tensorflow --clone ml_base
conda activate tensorflow
conda install -c anaconda tensorflow-gpu
python -m ipykernel install --user --name tensorflow --display-name "Python (tensorflow)"
conda deactivate
```

## 建立一個python Keras (基於Tensorflow)的conda環境

上述tensorflow環境必須先建製完成

```
conda create -n keras --clone tensorflow
conda activate keras
conda install -c conda-forge keras
python -m ipykernel install --user --name keras --display-name "Python (keras)" 
conda deactivate
```

## 建立一個python PyTorch的conda環境

上述ml\_base環境必須先建製完成

```
conda create -n pytorch --clone ml_base
conda activate pytorch
conda install -c pytorch pytorch 
python -m ipykernel install --user --name pytorch --display-name "Python (pytorch)"
conda deactivate
```

## jupyter kernel的管理

```
# check all your kernel
jupyter kernelspec list
# delete your kernel
jupyter kernelspec uninstall yourKernel
```