Turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Mute
- Printer Friendly Page

Highlighted
##

bingol__evren

Beginner

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

12-01-2019
12:49 PM

82 Views

Install prerequisite numpy version conflicts

I get this error

ERROR: tensorflow 1.15.0 has requirement numpy<2.0,>=1.16.0, but you'll have numpy 1.15.0 which is incompatible.

ERROR: mxnet 1.3.1 has requirement numpy<1.15.0,>=1.8.2, but you'll have numpy 1.15.0 which is incompatible.

When I upgrate to latest numby tensorflow error goes away

and I am only left with :

ERROR: mxnet 1.3.1 has requirement numpy<1.15.0,>=1.8.2, but you'll have numpy 1.15.0 which is incompatible.

WHen I down grade the the version mxnet wants tensorflow complains.

ERROR: tensorflow 1.15.0 has requirement numpy<2.0,>=1.16.0, but you'll have numpy 1.15.0 which is incompatible.

It is sort of a chicken and egg situation

Am I missing anything.

Thanks

1 Reply

Highlighted
##

Cary_P_Intel1

Employee

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

12-02-2019
12:38 AM

82 Views

Hi, Evren,

Yes, the required library for different framework might conflict with each other due to numpy version used. Do you use both of the framework at the same time? currently the workaround might need to change the numpy version base on the AI framework you are using or about to converted by Model Optimizer.

For more complete information about compiler optimizations, see our Optimization Notice.