We recommend using a cloud machine with high-performance GPUs for running these experiments. We use Hyperstack H100 for optimal performance. Other cloud options include:
- AWS EC2 (Deep Learning Base OSS Nvidia Driver GPU AMI - Ubuntu 24.04)
- Google Cloud Platform with A100/H100 GPUs
- Lambda Labs
- Paperspace
Run the setup script from server.sh to automatically install dependencies and configure the environment:
# Run the server.sh setup script
wget https://raw.githubusercontent.com/DDantalion/CSE232B/main/server.sh
sudo su
bash server.sh