Open darthmodel opened 1 year ago
Hi @YoshitakaMo, thank you for making localcolabfold available!
I also get the same error on our HPC system.
Computational environment
nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2021 NVIDIA Corporation Built on Mon_May__3_19:15:13_PDT_2021 Cuda compilation tools, release 11.3, V11.3.109 Build cuda_11.3.r11.3/compiler.29920130_0
Many thanks, Vasilis
Hi @vprobon,
Not sure if this helps, but I figured out the reason I was getting this is how my pip was configured. It wasn't pulling the packages from PyPi.
Cheers, Cui
Hi @darthmodel and thank you for the tip. I do not know if this applies to my case but I will try to find out ...
Best, Vasilis
What is your question? I'm new to
localcolabfold
. I installed it on an AWS instance with an Nvidia T4 GPU. The installation went smoothly. However, it complains about not being able to find TPU. I'm confused that why the code would look for a TPU, instead of the GPU there. Is there a way to force it to use the GPU? Thanks!Computational environment
G Cui