Installing previous versions of PyTorch

We’d prefer you install the latest version, but old binaries and installation instructions are provided below for your convenience.

Via conda

To install a previous version of PyTorch via Anaconda or Miniconda, replace “0.1.12” in the following commands with the desired version (i.e., “0.2.0”).

Installing with CUDA 8

conda install pytorch=0.1.12 cuda80 -c soumith

Installing with CUDA 7.5

conda install pytorch=0.1.12 cuda75 -c soumith

Installing without CUDA

conda install pytorch=0.1.12 -c soumith

From source

It is possible to checkout an older version of PyTorch and build it. You can list tags in PyTorch git repository with git tag and checkout a particular one (replace ‘0.1.9’ with the desired version) with

git checkout v0.1.9

Follow the install from source instructions in the of the PyTorch checkout.

Via pip

Download the whl file with the desired version from the list below, and run

pip install /path/to/whl/file

You can also directly pass a URL to pip:

pip install

CUDA support is optional; for example, you can install the cu80/torch-0.2.0.post3-cp36-cp36m-manylinux1_x86_64.whl package on a system that does not have CUDA.

There are three major lists below: Linux binaries compiled with CUDA 8 support, Linux binaries with CUDA 7.5 support, and Mac OSX & miscellaneous other binaries.

PyTorch Linux binaries compiled with CUDA 9.1

PyTorch Linux binaries compiled with CUDA 9.0

PyTorch Linux binaries compiled with CUDA 8

PyTorch Linux binaries compiled with CUDA 7.5

Mac and misc. binaries