Another "not a supported wheel on this platform" issue

1

Getting the “not a supported wheel on this platform” error when installing via pip on Ubuntu 14.04.5 64-bit with CUDA 8.0 on python 3.5.3
I’ve looked at the other issues relating to this and none offer solutions (I don’t want to use conda).

$ pip install -v http://download.pytorch.org/whl/cu80/torch-0.1.12.post2-cp35-cp35m-linux_x86_64.whl 
torch-0.1.12.post2-cp35-cp35m-linux_x86_64.whl is not a supported wheel on this platform.
Exception information:
Traceback (most recent call last):
  File "/home/jake/.virtualenvs/pytorch/lib/python3.5/site-packages/pip/basecommand.py", line 122, in main
    status = self.run(options, args)
  File "/home/jake/.virtualenvs/pytorch/lib/python3.5/site-packages/pip/commands/install.py", line 257, in run
    InstallRequirement.from_line(name, None))
  File "/home/jake/.virtualenvs/pytorch/lib/python3.5/site-packages/pip/req.py", line 167, in from_line
    raise UnsupportedWheel("%s is not a supported wheel on this platform." % wheel.filename)
pip.exceptions.UnsupportedWheel: torch-0.1.12.post2-cp35-cp35m-linux_x86_64.whl is not a supported wheel on this platform

$ uname -a
Linux hulk 3.13.0-119-generic #166-Ubuntu SMP Wed May 3 12:18:55 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux

$ lsb_release -a
No LSB modules are available.
Distributor ID:	Ubuntu
Description:	Ubuntu 14.04.5 LTS
Release:	14.04
Codename:	trusty

$ python --version
Python 3.5.3

$ nvcc --version                                                                                     
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2016 NVIDIA Corporation
Built on Tue_Jan_10_13:22:03_CST_2017
Cuda compilation tools, release 8.0, V8.0.61

$ pip --version
pip 1.5.4 from /home/jake/.virtualenvs/pytorch/lib/python3.5/site-packages (python 3.5)

$ virtualenv --version
1.11.4
2

On Debian and Ubuntu, you likely want pip3 instead of pip.

Best regards

Thomas

3

Tried this. Same output as with regular pip.

4

I must admit I didn’t use virtualenv but just sudo pip3’d it into localhost on Debian unstable. I try to avoid having numpy reinstalled, so I pass --no-deps, but the system pip3 liked the torch packages and put everything neatly under /usr/local/lib/python3.5.

Best regards

Thomas

5

Understandable, but I need it this way and I feel like it should work this way.

6

Upgrading pip solved this problem.