The notebooks can be run as is on the RHODS PyTorch notebook.
This demo requires a gpu with at least 24GB of memory. It also requires 2 buckets of S3 compatible storage.
- Create a data science project
- Open the OpenShift console
- Navigate to the created OpenShift project
- Import setup/setup-s3.yaml
- Create a pipeline server using the
Pipeline Artifacts
data connection. - Create a notebook using the PyTorch RHODS image and using the
My Storage
data connection - Clone this repo (
https://github.com/cfchase/text-to-image-demo.git
)