Installation on Linux via Conda

Anacondais a free and open source distribution of Python and R for computational science. Anaconda is dedicated to simplifying package management and deployment. Anaconda’s packages are managed using the package management system Conda. Conda is an open source package management system and environment management system that runs on Windows, macOS, and Linux.

Environmental preparation

Before performing PaddlePaddle installation, please make sure that your Anaconda software environment is properly installed. For software download and installation, see Anaconda’s official website ( If you have installed Anaconda correctly, follow these steps to install PaddlePaddle.

1.1 Create Virtual Environment

1.1.1 Create the Anaconda Virtual Environment

Create virtual environment First create the Anaconda virtual environment according to the specific Python version. The Anaconda installation of PaddlePaddle supports the following four Python installation environments.

If you want to use python version 3.6:

conda create -n paddle_env python=3.6

If you want to use python version 3.7:

conda create -n paddle_env python=3.7

If you want to use python version 3.8:

conda create -n paddle_env python=3.8

If you want to use python version 3.9:

conda create -n paddle_env python=3.9

1.1.2 Enter the Anaconda Virtual Environment

for Windows

activate paddle_env

for MacOS/Linux

conda activate paddle_env

1.2 Confirm Other Environments

Confirm that your conda virtual environment and the Python loaction which is preapared to install PaddlePaddle are where you expected them for your computer may have multiple Pythons environments. Enter Anaconda’s command line terminal and enter the following command to confirm the Python location.

1.2.1 Depending on your environment, you may need to replace python in all command lines in the instructions with specific Python path.

In a Windows environment, the command to get the Python path is:

where python

In a MacOS/Linux environment, the command to get the Python path is:

which python

1.2.2 Check the version of Python

Use the following command to confirm it’s version is 3.6/3.7/3.8/3.9

python --version

1.2.3 Confirm that Python and pip are 64bit, and the processor architecture is x86_64 (or x64, Intel 64, AMD64) architecture. Currently PaddlePaddle does not support arm64 architecture. The first line below print “64bit”, the second line prints “x86_64 (or x64, AMD64).”

python -c "import platform;print(platform.architecture()[0]);print(platform.machine())"


Choose CPU/GPU

  • If your computer doesn’t have NVIDIA® GPU, please install the CPU Version of PaddlePaddle

  • If your computer has NVIDIA® GPU, please make sure that the following conditions are met and install the GPU Version of PaddlePaddle

    • CUDA toolkit 10.1/10.2 with cuDNN 7 (cuDNN version>=7.6.5, for multi card support, NCCL2.7 or higher)

    • CUDA toolkit 11.2 with cuDNN v8.1.1(for multi card support, NCCL2.7 or higher)

    • Hardware devices with GPU computing power over 3.5

      You can refer to NVIDIA official documents for installation process and configuration method of CUDA and cudnn. Please refer to CUDAcuDNN

  • If you need to use a multi-card environment, please make sure that you have installed nccl2 correctly, or install nccl2 according to the following instructions (here are the installation instructions of nccl2 under CUDA10.2 and cuDNN7. For more version installation information, please refer to NVIDIA Official Website):

    • Centos system can refer to the following commands

      rpm -i nvidia-machine-learning-repo-rhel7-1.0.0-1.x86_64.rpm
      yum update -y
      yum install -y libnccl-2.7.8-1+cuda10.2 libnccl-devel-2.7.8-1+cuda10.2 libnccl-static-2.7.8-1+cuda10.2
    • Ubuntu system can refer to the following commands

      dpkg -i nvidia-machine-learning-repo-ubuntu1604_1.0.0-1_amd64.deb
      sudo apt install -y libnccl2=2.7.8-1+cuda10.2 libnccl-dev=2.7.8-1+cuda10.2

Installation Step

You can choose the following version of PaddlePaddle to start installation:

2.1 CPU Version of PaddlePaddle

conda install paddlepaddle==2.1.3 --channel

2.2 GPU Version of PaddlePaddle

  • If you are using CUDA 10.1,cuDNN 7 (cuDNN version>=7.6.5):

conda install paddlepaddle-gpu==2.1.3 cudatoolkit=10.1 --channel
  • If you are usingCUDA 10.2,cuDNN 7 (cuDNN version>=7.6.5):

conda install paddlepaddle-gpu==2.1.3 cudatoolkit=10.2 --channel
  • If you are using CUDA 11.2,cuDNN 8 (cuDNN version>=8.1.1):

conda install paddlepaddle-gpu==2.1.3 cudatoolkit=11.2 -c -c conda-forge

Verify installation

After the installation is complete, you can use python or python3 to enter the Python interpreter and then use import paddle and paddle.utils.run_check()

If PaddlePaddle is installed successfully! appears, to verify that the installation was successful.


For domestic users who cannot connect to the Anaconda official source, you can add Tsinghua source according to the following command.

conda config --add channels
conda config --add channels
conda config --set show_channel_urls yes