-
Notifications
You must be signed in to change notification settings - Fork 0
/
slurmTemplate20.sh
42 lines (33 loc) · 1.34 KB
/
slurmTemplate20.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
ww721@ese-hivemind:/raid/hivemind$ cat slurmTemplate20.sh
#!/bin/bash -l
##############################
# Job blueprint #
##############################
# Give your job a name, so you can recognize it in the queue overview
# e.g. The jobname below is "example"
#SBATCH --job-name=example
# Each job will utilise all of brainstorm's resources.
# Here, we ask for 1 node with exlusive
# brainstorm has 256 CPU cores and 8 A100 GPUs.
####--cpus-per-task 8 --gres=gpu:1 --mem-per-cpu 1500 --time 24:00:00 --pty bash -i
#SBATCH -e slurm-%j.err # File to redirect stderr
#SBATCH -o slurm-%j.out # File to redirect stdout
#SBATCH --mem-per-cpu=1500 # Memory per processor
#SBATCH --time=0-00:04:00 # The walltime
#SBATCH --nodes=1 # Run all processes on a single node
#SBATCH --ntasks=2 # Number of tasks
#SBATCH --ntasks-per-socket=1 # Maximum number of tasks on each socket
#SBATCH --cpus-per-task=5 # Number of CPU cores per task
#SBATCH --gres=gpu:2 # Number of GPUs to allocate to this job
# This is where the actual work is done.
#nvidia-smi
#nvcc --version
source ~/miniconda3/etc/profile.d/conda.sh
conda activate testgpu
#numba -s
# These are individual tasks
python xavier.py 0&
python xavier.py 1&
wait
# Finish the script
exit 0