Tensorflow GPU causing fatal error

[sanseveroj]

Hello I am new to posting here so I will do my best to explain my situation if I need to add any details please let me know what the standard procedure is!

I have created a local conda environment with python 3.7 and have successfully installed tensorflow-gpu to it. Yet when I run: tf <- import('tensorflow') in R I always get the following output:

2020-02-29 08:24:54.101489: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_101.dll

Then I immediately get an error saying "R session aborted R encountered a fatal error..." I'm certain we are all familiar with the screen. I was wondering if anyone might have some tips for diagnosing the problem.

I have a nvidia 2080ti, windows 10, and all of the up to date software supporting tensorflow (I.E. Py 3.7, R 3.6.2)

Again let me know if there is any additional information I can provide.

[dfalbel]

Can you try running this from the command line so we can see the full stack trace?
On windows you may be able to do it by following the steps here: http://datacornering.com/how-to-run-r-scripts-from-the-windows-command-line-cmd/