Différences

Ci-dessous, les différences entre deux révisions de la page.

Lien vers cette vue comparative

Les deux révisions précédentes Révision précédente
Prochaine révision
Révision précédente
deeplearning:dellvolta100 [2020/11/05 12:23] – [Anaconda] admindeeplearning:dellvolta100 [2022/01/11 16:30] (Version actuelle) – [Using default installed software] admin
Ligne 167: Ligne 167:
    
 ===== Using default installed software ===== ===== Using default installed software =====
 +<note>Please use anaconda for installing packages.</note> 
 ==== Vanilla Python ==== ==== Vanilla Python ====
  
Ligne 219: Ligne 220:
 </code> </code>
 */ */
-==== Anaconda ====+==== Anaconda [Recommended] ====
  
 For better performances Anaconda python V3 is installed in ''/opt/''. Several optimized scientific packages are available.  For better performances Anaconda python V3 is installed in ''/opt/''. Several optimized scientific packages are available. 
  
  
-/* 
-<note important>Since ''$HOME'' directory is read only on all computing nodes.\\ 
-We need to change the default location of ''.conda'' directory. 
-  * Create ''.conda'' folder in your ''$WORK'' directory <code bash>$ mkdir -p $WORK/.conda</code> 
-  * Delete/move old ''$HOME/.conda'' config if exist <code bash>$ mv $HOME/.conda $HOME/.conda.save  </code> 
-  * Make a symbolic link from the folder to your ''$HOME'' <code bash>$ ln -s  $WORK/.conda $HOME/.conda </code>   
- 
-</note> 
-*/ 
- 
-<note important>Since ''$HOME'' directory is read only on all computing nodes.\\ 
-We need to make ''$WORK'' as Home directory: 
- 
-<code bash> 
-$export HOME=$WORK 
-</code> 
-</note> 
  
  
Ligne 255: Ligne 239:
 <note tip>**EDIT (10/03/2020)**: we added ''module'' command only for local software in these nodes. <note tip>**EDIT (10/03/2020)**: we added ''module'' command only for local software in these nodes.
 To load anaconda simply use: To load anaconda simply use:
-<code bash>$ module load anacoda</code></note>+<code bash>$ module load anaconda</code></note>
  
  
Ligne 291: Ligne 275:
 You may need to install additional package without **having** root access. We instead recommend building a ''virtual environment'' locally and install packages there.\\ You may need to install additional package without **having** root access. We instead recommend building a ''virtual environment'' locally and install packages there.\\
 But before we need to tell conda to use WORK to store pkgs : But before we need to tell conda to use WORK to store pkgs :
-<code bash>conda config --add pkgs_dirs $WORK/.conda/pkgs</code>+''export HOME=$WORK''
  
-  - Create a conda environment <code bash>$ conda create -y --prefix WORK/conda-env</code> replace ''conda-env'' by a valid name. +  - Create a conda environment <code bash>$ conda create -y --prefix $WORK/conda-env</code> replace ''conda-env'' by a valid name. 
   - Activate the environment you just created <code bash>$ source activate $WORK/conda-env  </code>   - Activate the environment you just created <code bash>$ source activate $WORK/conda-env  </code>
   - Download and install the conda package in the environment <code bash> $ conda install package</code> replace package by the package you wan to install, for example ''panda''   - Download and install the conda package in the environment <code bash> $ conda install package</code> replace package by the package you wan to install, for example ''panda''
Ligne 361: Ligne 345:
 */   */  
 ===== Using Singularity v3 ===== ===== Using Singularity v3 =====
 +<note warning>This section is for exprerts and advanced usage</note>
 We can use Singularity's ''pull'' sub-command to import a container image directly from [[https://hub.docker.com|Docker Hub]]  without having root or superuser privileges (or Docker) on your host system. We can use Singularity's ''pull'' sub-command to import a container image directly from [[https://hub.docker.com|Docker Hub]]  without having root or superuser privileges (or Docker) on your host system.
  
  • deeplearning/dellvolta100.1604579028.txt.gz
  • Dernière modification : 2020/11/05 12:23
  • de admin