Jupyter notebooks in the NVIDIA cuQuantum Appliance #33
Answered
by
zohimchandani
zohimchandani
asked this question in
Q&A
-
|
I pulled the image from here and started a container. In the following directory: Multiple kernels are available to me in here as shown in the image below:
Why is the container shipped with multiple versions of python? When I try run the code snippet from here), with the conda env shown in the image above, it asks me to install ipykernel which I do as shown below: This I believe changes the path for How do I circumvent this to make the appliance support Thanks |
Beta Was this translation helpful? Give feedback.
Answered by
zohimchandani
Jan 27, 2023
Replies: 4 comments 1 reply
-
|
Note that the location of |
Beta Was this translation helpful? Give feedback.
0 replies
-
|
Convert this to a discussion thread since it's not a bug. |
Beta Was this translation helpful? Give feedback.
0 replies
-
|
conda is installing its own mpi library package and overriding the system's CUDA-enabled mpirun. and set the environment variables in the conda environment before running qiskit: |
Beta Was this translation helpful? Give feedback.
0 replies
-
|
When creating a Following the suggestion above would change the path of Instead, pick the |
Beta Was this translation helpful? Give feedback.
1 reply
-
|
Thanks, @zohimchandani. Indeed, the correct Conda environment ( root@850229efe3af:/workspace# source /opt/conda/bin/activate
(base) root@850229efe3af: conda install -c conda-forge ...We'll check if we could activate it automatically for users, if so we'd fix it in the next release (internal ticket CUQNT-1676). |
Beta Was this translation helpful? Give feedback.
Answer selected by
zohimchandani
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
When creating a
.ipynbfile in theexamplesfolder, thePython 3.8.10kernel in the/bin/python3directory raises the following error:Following the suggestion above would change the path of
mpiruncausing issues further down.Instead, pick the
base (Python 3.8.15)kernel inopt/conda/bin/pythondirectory which haspipinstalled and dopip3 install ipykernelto be able to use notebooks in the container.