run_character_on_cluster.sh

#!/bin/bash
#SBATCH --job-name=character_patchloss_for_ganwriting
#SBATCH --ntasks=1
#SBATCH --cpus-per-task=4
#SBATCH --mem=4000
#SBATCH --gres=gpu:1
#SBATCH -o /home/%u/%x-%j-on-%N.out
#SBATCH -e /home/%u/%x-%j-on-%N.err
#SBATCH --mail-type=ALL
#SBATCH --time=24:00:00

export WORKON_HOME==/cluster/`whoami`/.python_cache

echo "Your job is running on" $(hostname)
pip3 install --user python-Levenshtein
# copy to /scratch/qi69dube
mkdir /scratch/qi69dube
cp -nr /cluster/qi69dube/dataset.tar.gz /scratch/qi69dube/dataset.tar.gz
tar xfk /scratch/qi69dube/dataset.tar.gz -C /scratch/qi69dube/
# -1 stands for: continue with the last saved-model
python3 main_run.py -1 save_weights_character imgs_character --character_patch_loss