Intel® Optimized AI Frameworks
Get community support for questions related to PyTorch* and TensorFlow* frameworks.
Announcements
Welcome to the Intel Community. If you get an answer you like, please mark it as an Accepted Solution to help others. Thank you!

Can't upgrade tensorflow

Praburam__Myleshen
346 Views

Whenever i try to update tensorflow i'm getting memory error in my login node(The error is : "Process exceeded login node resource limit" ), should i update all the modules in my login node or in the computation node using qsub-I 

The command which is used is 

     pip install tensorflow --user --upgrade

 

And also is there any tutorial for how to install modules in Devcloud cause the original tutorial (documentation) does not contain that much of information

0 Kudos
3 Replies
AthiraM_Intel
Moderator
346 Views

Hi,

Thanks for reaching out to us.

You can update all the modules in the compute node. In most cases, memory error is caused by trying to run compute-intensive tasks on the login node. In such cases, login to the compute node using qsub –I and execute your commands there.

For installing modules using pip you can follow the original documentations. Run pip with the argument –user to install the same in DevCloud.

Please refer https://readthedocs.org/projects/devcloud-docs/downloads/pdf/latest/ for more information on using DevCloud

Hope this clarifies your query.

 

Praburam__Myleshen
346 Views
Yeah, thanks a lot for the clarification. My doubt is cleared now. But I would like you guys to make the new documentation easily available and some basic tutorials in youtube if possible, and yeah you can close this thread
AthiraM_Intel
Moderator
346 Views

Hi,

Thank you for the confirmation.

We do have some online courses available for machine learning and deep learning.You can have a look at them:

https://software.intel.com/en-us/ai/courses/deep-learning

https://software.intel.com/en-us/ai/courses/machine-learning

Since your initial query is resolved,we are closing this case.Please feel free to open a new thread if you face any further queries.

Thanks.

Reply