while i am trying to train models for nvidia tao with google colab, i always struggle because Google makes some updates in colab and the notebooks start to fail in some parts. The last one was about the python version again.
Actually it was solved by nvidia developers by updating .sh of colab in nvidia-tao repo. However its again started and i spend hours of trial for solving the issue.
I tried Google Cloud Platform and Amazon AWS but in that position i am getting problems to install dependencies and find the required gpu machine.
If you know a way to create a stable environment I will train my model using notebooks please help me by recommending any tutorial.
1 post - 1 participant