Closed StevenJokess closed 4 years ago
Welcome to Apache MXNet (incubating)! We are on a mission to democratize AI, and we are glad that you are contributing to it by opening this issue. Please make sure to include all the relevant context, and one of the @apache/mxnet-committers will be here shortly. If you are interested in contributing to our project, let us know! Also, be sure to check out our guide on contributing to MXNet and our development guides wiki.
Reproduced the issue. As on linux upgrading pip to newest release helped, on win10 was unable to see the package even with pip==20.2.4.
@szha Do you know what may be the cause?
I think the problem was that the windows pip package for mxnet has become too large to publish on PyPI.
@szha I understand. I tried looking for this package but only found whl file for mxnet==1.7.0 (not 1.7.0.post1) for windows. Is this package (1.7.0.post1 for win10) available anywhere?
they should have the same content
@StevenJokess Try this package: https://repo.mxnet.io/dist/python/cpu/mxnet-1.7.0-py2.py3-none-win_amd64.whl. Please let me know if it works for you.
@bartekkuncer Where did you find this package?
@StevenJokess On internal CI page. Does it work? :)
Thanks. How to find it next time? Or another wheel?
->How to find it next time? Or another wheel? If the package will not decrease in size for future releases I believe it will be hard to publish them on PyPl. @szha Is there any other public channel to get the win10 version of 1.7.0 and other mxnet release packages?
Not that I'm aware of. @yajiedesign was looking into custom compression that may help this, though I think a more reliable way forward is for NVIDIA to publish these dependencies as a separate package that we can depend on.
Let's track this in #18793 which was intended for tracking CD automation for windows.
@StevenJokess I believe as of now there is no way to find those packages for windows on your own. You can always try building mxnet from source to have recent versions: https://mxnet.apache.org/versions/1.7.0/get_started/build_from_source.
I would also like to pip install mxnet==1.7.0 on windows. @szha How sure are you about the size issue mentioned above? I noticed that two other mxnet whl files (for macos and linux) are already on PyPi and they are much larger than the one that @bartekkuncer helpfully referenced on Monday https://github.com/apache/incubator-mxnet/issues/19594#issuecomment-735826843 . That whl file was only about 17MB. The macos and linux whl files are 30MB and 55MB.
I wonder if someone could again try to send the Windows whl file to PyPi? It would be a big help to me. Thanks!
the windows version should be available on pypi now
hi, same problem here
I tried to activate the package of @bartekkuncer but not working either...
can anyone help me?
@dmteams what you have seems to be a different problem where compilation for numpy failed. you can try to first install numpy directly (to hopefully pick up pre-built binary instead of building from source) and then install mxnet.
@dmteams I see that the numpy package was taken from cache. It is possible that there were some problems during download and therefore it does not work now. You might want to try pip install --no-cache-dir mxnet
which will ignore cached libs. Please let me know if it helps.
If it goes to missing environment, you have to create one first to be able to activate it. I believe you do not need it to install mxnet with pip.
I tried to uninstall/reinstall numpy successfully, then try both pip install mxnet
and pip install --no-cache-dir mxnet
but I still have the same problem. I really do not understand... I finally installed TensorFlow that seems working just fine... Thank you anyway guys!
Description
(A clear and concise description of what the bug is.)
Error Message
(Paste the complete error message. Please also include stack trace by setting environment variable
DMLC_LOG_STACK_TRACE_DEPTH=100
before running your script.)To Reproduce
(If you developed your own code, please provide a short script that reproduces the error. For existing examples, please provide link.)
Steps to reproduce
(Paste the commands you ran that produced the error.)
1.
What have you tried to solve it?
Environment
We recommend using our script for collecting the diagnostic information with the following command
curl --retry 10 -s https://raw.githubusercontent.com/apache/incubator-mxnet/master/tools/diagnose.py | python3
Environment Information
``` # Paste the diagnose.py command output here ```How to download mxnet==1.7.0.post1?