Deep Learning Course

Ressources for Deep Learning Course (VSE).

GitHub Repository:

https://github.com/MICS-Lab/DeepLearningCourse2025

Grade breakdown:

Drive folder with datasets:

https://drive.google.com/drive/folders/14QmpxUPmpeRtUxnDiCIaNLlsx9KIv8AK?usp=sharing

Kaggle competition links:

• Fitting 1D function: https://www.kaggle.com/competitions/2025-fitting-a-1-d-1-d-function-with-deep-learning
• MICS MNIST: https://www.kaggle.com/competitions/2025-decimal-classification-mics-mnist

Linux operations to install Python on the server at your disposal for GPUs

#!/bin/bash

#SBATCH --job-name=condaEnvSetup
#SBATCH --nodes=1
#SBATCH --partition=gpu_prod_long
#SBATCH --time=1:00:00

# Ensure module is visible
source /etc/profile

# Load the conda module
module load anaconda3/2022.10/gcc-13.1.0

conda remove -n DLNLP2025 --all

# Create conda environment
# and prevent the use of system site-package
# that will interfer with the conda env
export PYTHONNOUSERSITE=1

# Create the environment using python 3.12
conda create --name DLNLP2025 python=3.12 --force

# Activate the environment
source activate DLNLP2025

pip install torch torchvision torchaudio matplotlib numpy pandas seaborn scipy scikit-learn ipykernel

Run a Python file

#!/bin/bash

#SBATCH --job-name=condaEnvSetup
#SBATCH --nodes=1
#SBATCH --partition=gpu_prod_long
#SBATCH --time=24:00:00

# Ensure module is visible
source /etc/profile

# Load the conda module
module load anaconda3/2022.10/gcc-13.1.0

# Activate the environment
source activate DLNLP2025

python my_file.py