json2tfrecord: write slurm job file

This commit is contained in:
Starbeamrainbowlabs 2022-08-08 15:53:32 +01:00
parent 222a6146ec
commit f6f2e3694c
Signed by: sbrl
GPG key ID: 1BE5172E637709C2

View file

@ -0,0 +1,41 @@
#!/usr/bin/env bash
#SBATCH -J Json2TfR
#SBATCH -N 1
#SBATCH -n 28
#SBATCH -o %j.%N.%a.out
#SBATCH -e %j.%N.%a.err
#SBATCH -p gpu05,gpu
#SBATCH --time=3-00:00:00
module load utilities/multi
module load readline/7.0
module load gcc/10.2.0
module load cuda/11.5.0
module load python/anaconda/4.6/miniconda/3.7
INPUT="${INPUT:-$HOME/rainfallwater_records}";
OUTPUT="${OUTPUT:-$HOME/rainfallwater_records_tfrecord}";
if [[ -z "${INPUT}" ]]; then
echo "Error: No input directory specified in the INPUT environment variable.";
exit 1;
fi
if [[ -z "${OUTPUT}" ]]; then
echo "Error: No output directory specified in the OUTPUT environment variable.";
exit 1;
fi
if [[ ! -d "${INPUT}" ]]; then
echo "Error: That input directory either doesn't exist, isn't a directory, or we don't have permission to access it.";
exit 3;
fi
export PATH=$HOME/software/bin:$PATH;
echo ">>> Settings";
echo "INPUT $INPUT";
echo ">>> Converting dataset .jsonl.gz → .tfrecord.gz";
/usr/bin/env time -v src/index.mjs jsonl2tfrecord --verbose --source "${INPUT}" --target "${OUTPUT}";
echo ">>> exited with code $?";