Versions Compared

Version Old Version 25 New Version 26
Changes made by

tianche5

tianche5

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.
Panel
panelIconId1f40d
panelIcon:snake:
panelIconText🐍
bgColor#E3FCEF

This example is about installing and using two Python packages pytorch and multiqc on Sol.

Table of Contents
minLevel1
maxLevel6
outlinefalse
stylenone
typelist
printabletrue

Pytorch

Go to https://pytorch.org/get-started/locally/ and select options as shown below:

...

Pytorch is special, the correct command should always be gathered from this webpage (not anaconda.org). Other packages may also have such conflicts, when encountered, make sure to follow the instructions on the package official website or github repo.

...

Multiqc

Go to anaconda.org and search for multiqc in the search bar:

...

The “conda install” on the above page gives the installation commands:

...

The conda part in the red circle needs to be changed to mamba, and the bioconda part in the blue box is the channel name of this multiqc package. Channel is similar to the name of an online folder, that mamba can find and download the correct package. This is important for the next step.

Pytorch

Go to https://pytorch.org/get-started/locally/ and select options as shown below:

...

Pytorch is special, the correct command should always be gathered from this webpage (not http://anaconda.org ). Other packages may also have such conflicts, when encountered, make sure to follow the instructions on the package official website or github repo.

The command given here CANNOT be directly used on ASU supercomputers, the sections below will show you how to re-write an install command.

Step 2 - Install

  1. Connect to the VPN

  2. Open a command line interface on Sol by navigating to sol.asu.edu on your browser and
    selecting the "Sol Shell Access" option from the "System" menu option. Or by SSHing into
    Sol using the command ssh <asurite>@sol.asu.edu

  3. interactive -p htc -c 4 -t 30

  4. module load mamba/latest

  5. module load cuda-12.4.1-gcc-12.1.0

  6. Here we combine three things into a single command:

    1. Create an environment

    2. Install Pytorch using the command found in step 1

    3. Install Multiqc using the command found in step 1

      Code Block
      mamba create -n myENV -c conda-forge -c pytorch -c nvidia -c bioconda python=3 multiqc pytorch torchvision torchaudio pytorch-cuda=12.4

  7. And answer Y to the promoted question, if everything looks fine

  8. Wait till the installation finish

...