PyTorch Installation

1 PyTorch

PyTorch is a popular open-source deep learning framework widely used for research and production applications. Developed by Facebook’s AI Research lab, PyTorch offers a dynamic computational graph, making it a flexible choice for deep learning tasks. It provides an extensive collection of tools and libraries for building and training neural networks.

2 PyTorch Software Environment Setup

It is recommended to use Conda to install PyTorch package. For detailed instructions on how to use Conda, please refer to the following link for instructions: https://www.carc.usc.edu/user-guides/data-science/building-conda-environment

2.1 Step 1: Request for an interactive Session

[user@discovery1 ~]$ salloc --partition=gpu --gres=gpu:1 --cpus-per-task=8 --mem=32GB --time=1:00:00 
salloc: Pending job allocation 15731446
salloc: job 15731446 queued and waiting for resources
salloc: job 15731446 has been allocated resources
salloc: Granted job allocation 15731446
salloc: Waiting for resource configuration
salloc: Nodes a02-15 are ready for job
[user@a02-15 ~]$ 

2.2 Step 2: Load Conda module and initialize shell to use Conda and Mamba (Step 2 is only needed if it is the first time you are using Conda on the cluster)

Note: Mamba is a drop-in replacement for most conda commands that enables faster package solving, downloading, and installing

module purge
module load conda
mamba init bash
source ~/.bashrc
module purge

2.3 Step 3: Create a virtual environment & install PyTorch package

After you finish initialization of shell in Step 2, you can start the PyTorch installation process using Conda with the following commands:

mamba create --name torch-env
mamba activate torch-env
mamba install pytorch torchvision torchaudio pytorch-cuda=11.8 -c pytorch -c nvidia