It's not the same cuda version on GPU 10 and the rest, I think different version of tensorflow has to be installed.<div><br></div><div>Check your tensorflow version and if it supports the cuda version on gpu10.</div><div><br></div><div><br><br>On Saturday, March 9, 2019, Predrag Punosevac <<a href="mailto:predragp@andrew.cmu.edu">predragp@andrew.cmu.edu</a>> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="auto">Try CUDA 10.0 instead of 10.1</div><div class="gmail_extra"><br><div class="gmail_quote">On Mar 9, 2019 5:28 PM, Yichong Xu <<a href="mailto:yichongx@cs.cmu.edu" target="_blank">yichongx@cs.cmu.edu</a>> wrote:<br type="attribution"><blockquote style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="auto">
Same issue here.<br>
<br>
<div dir="ltr">From my iPhone</div>
<div dir="ltr"><br>
On Mar 9, 2019, at 4:01 PM, Emre Yolcu <<a href="mailto:eyolcu@andrew.cmu.edu" target="_blank">eyolcu@andrew.cmu.edu</a>> wrote:<br>
<br>
</div>
<blockquote>
<div dir="ltr">
<div>
<p>Hi,</p>
<p> </p>
<p>Right now on gpu10 `nvcc --version` and `nvidia-smi` seem to work, but `python -c ‘import torch; print(torch.cuda.is_available(<wbr>))’` prints False. Is anybody running into the same issue?</p>
<p> </p>
<p>Emre</p>
</div>
</div>
</blockquote>
</div>
</blockquote></div><br></div></blockquote></div>