We from PyTorch Release Engineering team are actively working on PyTorch 2.0 release and wish to call out the support change concerning Python 3.11 support for Anaconda. Due to lack of Python 3.11 support for packages that PyTorch depend on, including NumPy, SciPy, SymPy, Pillow and others on Anaconda platform. We will not be releasing Conda binaries compiled with Python 3.11 for PyTorch Release 2.0. The Pip packages with Python 3.11 support will released, hence if you intent to use PyTorch 2.0 with Python 3.11 please use our Pip packages. You can download them from our test channel by using:
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/test/cpu
PIP CUDA 11.7, 11.8
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/test/cu117
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/test/cu118
Please note: Conda packages with Python 3.11 support will be made available on our nightly channel. Also we are planning on releasing Conda Python 3.11 binaries as part of Pytorch minor release 2.0.1 once Anaconda provides these key dependencies.
Please refer to this anaconda blog for more information on Anaconda releases:
Please refer to Release Compatibility Matrix here:
If you have any questions please comment below or reach out to us on this thread.