'Installing tensorflow on virtualbox ubuntu 20.04 python 2.7 - 'Illegal instruction (core dumped)'
my goal is to follow this guide so that I can convert a Caffe model to a Tensorflow model. As my original OS is Windows 10 I am using the virtual Ubuntu 20.04 (using Oracle VirtualBox) with python 2.7 and anaconda virtual env. I am able to successfully install tensorflow for python 2.7, but when I run the command python -c 'import tensorflow', I get an error with text Illegal instruction (core dumped). I have tried to google to get more information but all I got was that it maybe has something to do with the architecture, I found out that my architecture (on the virtual ubuntu) is x86_64. That is all I have found out and now I turned to SO to ask what I should do to be able to run tensorflow in the before described environment, many thanks beforehand.
Solution 1:[1]
For those wondering, the pip or conda installation did not work in my case, so I followed the instructions on this guide for linux, python 2.7 using the pip installation method with a tensorflow version of 0.10, which for my purposes is good enough.
Exact steps in case the link does not work in the future:
$ conda create -n tensorflow python=2.7$ conda activate tensorflow- Your prompt should change to this:
(tensorflow)$ (tensorflow)$ export TF_BINARY_URL=https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.10.0-cp27-none-linux_x86_64.whl(tensorflow)$ pip install --ignore-installed --upgrade $TF_BINARY_URL
This is, as outlined, for python 2.7, but the guide contains info for newer versions as well. To check whether tensorflow installed correctly you can run this command $ python -c 'import tensorflow as tf; print(tf.__version__)' which should output 0.10.0.
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
| Solution | Source |
|---|---|
| Solution 1 | Michael Kro?ka |
