3

I tried to run a code on my pc but I got

RuntimeError: CUDA out of memory. Tried to allocate 2.00 MiB (GPU 0; 4.00 GiB total capacity; 2.72 GiB already allocated; 1.83 MiB free; 4.76 MiB cached)

For this reason, I decided to run it on Google Colab, but I need to install all the requirements, one of these is tensorflow 1.12.2, and when I try to execute

!pip install tensorflow==1.12.2

I get

ERROR: Could not find a version that satisfies the requirement tensorflow==1.12.2

I have seen that in my Anaconda environment, I use python 3.6.13, while in Google Colab is present python 3.7.10. Could it be that the problem? I tried to downgrade the python version in Google Colab, but I was not able to downgrade it. Could anyone help me, please?

GooDeeJAY
  • 1,681
  • 2
  • 20
  • 27
Azzurra
  • 51
  • 6
  • 2
    https://stackoverflow.com/a/62855139/12652702 Hope this helps – Rahil Kadakia Apr 27 '21 at 09:40
  • Yes, due to python incompatibility you received above error. Are there any specific reason that you want to use `TF 1.12.2` ? You can choose required version by connect to a [local runtime](https://research.google.com/colaboratory/local-runtimes.html). Thanks! –  May 11 '21 at 14:24

0 Answers0