Create GPU kernel.txt

conda create -n gpu2 python=3.9
conda activate gpu2
pip install --upgrade pip
pip install tensorflow
pip install nvidia-cudnn-cu11==8.6.0.163
pip install torch
pip install torchvision

test if build with cuda using :
python 
>> import tensorflow as tf
>> tf.test.is_built_with_cuda()
--> should return True

conda install -c anaconda ipykernel

python -m ipykernel install --user --name=gpu2 #create the kernel into the juputer notebook

--> open a notebook 

import tensorflow as tf
from tensorflow.python.client import device_lib
tf.config.list_physical_devices('GPU')

--> should return something like [PhysicalDevice(name='name', device_type = 'GPU']

device_lib.list_local_devices()

--> should see your GPU : i.e. NVIDIA etc....




cd /mnt/c/Users/strom/Desktop/fmri_project/subjects/sub-01

#!/bin/bash
#SBATCH --chdir /scratch/izar/<put-your-username-here>
#SBATCH --nodes 1
#SBATCH --ntasks 1
#SBATCH --cpus-per-task 1
#SBATCH --mem 8G
#SBATCH --partition gpu
#SBATCH --gres gpu:1
#SBATCH --qos dlav
#SBATCH --account civil-459-2023
conda activate PyTorch
python main_train.py