Use the file pJupyter.batch inside the Use_jupyter_notebook, you could modify:
The partition you are using:
#SBATCH --partition=omicsbio
The number of tasks
#SBATCH --ntasks=1
The memory size you want to use:
#SBATCH --mem=10G
The node you would like to use
#SBATCH --nodelist=c660
Time to run:
#SBATCH --time=12:00:00
Working directory
#SBATCH --chdir=/work/omicsbio/user007/
The port use for the connection:
port=8893
Load the modules that you'll use
e.g.
module load Python/3.6.4-foss-2018a
module load TensorFlow/1.8.0-foss-2018a-Python-3.6.4
In order to connect the jupyter notebook to your local machine, it is neccesary to create a tunnel
In Putty:
It would be necessary to configurate a tunnel. Go to Connection/SSH/Tunnels. If you didn't specify a node, you could use the command squeue or see the output file (jupyter_notebook_number.log). For instance, with node 660 and port 8893, fill the options in this way, then press Add.
After you submit your job
sbatch pJupyter.batch
Go to the output jupyter_notebook_number.log (e.g. jupyter_notebook_34036888.log) and look for a line with the following syntax:
e.g.
http://127.0.0.1:8893/?token=48fc02e6eaef0756fdbcf6e7767336620be5c84afccd45f9
Copy and paste in your browser. Now you are ready to use jupyter.