Closed kwesiRutledge closed 1 year ago
FWIW I was able to reproduce this on my machine by upgrading from torch==1.9.1
to torch==1.13.1
.
Just stumbled upon this while cleaning out old emails.
My apologies, if this was an error that occurred due to me accidentally upgrading torch. I thought that I executed my instructions from a fresh install, but it looks like I had torch==1.12.1
in the requirements.txt for some reason. I'll close this issue now.
(This issue is described partially in Pull Request #10 .)
Steps To Reproduce
The following steps were executed on MacOS 13.0.1 (22A400). Note that, following Apple's directions, the Conda version that was used in this installation is Miniconda (Inspired by this tutorial.)
Attempt to follow the installation steps provided in the project README. These instructions failed. To successfully complete the step
pip install -r requirements.txt
, I needed to:conda install casadi
and thenpip install diffcp==1.0.18
.Once everything is installed, change your directory to
neural_clbf > neural_clbf > training
. Run the scripttrain_inverted_pendulum.py
. Before the first Epoch starts, you will receive the following error and the script will terminate.Error
Full Stacktrace