Skip to content

This is wrapper script for submitted python scripts to the CLSP grid

Notifications You must be signed in to change notification settings

sonal-ssj/clsp_grid_wrapper

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Authors

This watered down script is based on Raghavendra Reddy Pappagari's submission scripts

Usage

submit_grid.sh <YOUR_PYTHON_SCRIPT>

For example, to submit a python script called test_pytorch.py to the grid, run submit_grid.sh test_pytorch.py This will activate the virtual environment as in pytorch_run.sh and run a script to check if gpu is being used with pytorch.

Files and description

.
├── `submit_grid.sh` : Script to submit a python script to the grid
├── `pytorch_run.sh` : Sample script that activates virtual environment with pytorch and gets CUDA stuff
└── `test_pytorch.py` : Sample script to check if pytorch is installed and uses GPU

Optional: You can have multiple scripts like tensorflow_run.sh : Script that activates virtual environment with tensorflow and gets CUDA stuff keras_run.sh : Script that activates virtual environment with keras and gets CUDA stuff

Sanity check run

  1. Make sure your python and conda path is set in ~/.bashrc. Run

    source ~/.bashrc
    which python
    which conda
    

    This should give the python and conda path you desire to use

  2. Edit the name of your virtual environment which has pytorch in pytorch_run.sh. This can be done by replacing <YOUR_ENV_NAME> in line 2 of pytorch_run.sh by something like my_env. Now, the pytorch_run.sh file must read like

    # Activate your virtual environment
    conda activate my_env
    
    CUDA_VISIBLE_DEVICES=`/home/gkumar/scripts/free-gpu` python  "$@"
    
  3. Edit submit_grid.sh to change the names of log and output files, else by default they will be named by using TIMESTAMP. You may also have to change memory requirements in line 26 (mem_free, ram_free)as per your script's requirement.

    # qsub command, you can edit the memory requirements
    export qsub_cmd="qsub -cwd -l mem_free=8G,ram_free=8G,gpu=1,hostname=b1[123456789]|c* -q g.q"
    
  4. Give execute permission to submit_grid.sh by running:

    chmod +x submit_grid.sh 
    
  5. Run submit_grid.sh test_pytorch.py

  6. You should get output like follows in *.out file.

    Pytorch available?
    True
    Cuda available?
    True
    Success!
    

About

This is wrapper script for submitted python scripts to the CLSP grid

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published