-
In the My Projects, click Create Project.
Project in SLM-UI is a logical representation of a particular Swarm training.
Projects help to view deployment topology and monitor the progress for the given Swarm training. They define what all Swarm nodes (and associated host nodes) a training will run, the model being used, the x.509 certificates, SWOP and Task yaml files for a particular training. Multiple Projects can be defined in a single instance of SLM-UI.
Project artifacts are created under the swarm-learning/slm-ui/projects/<project number> automatically once the project is saved.
NOTE:
HPE recommends that users avoid editing these files directly on the host machine.
-
Enter the project name.
-
Select the Swarm version number from the drop-down menu.
-
Upload the following files:
-
Upload your Certificate and CA Certificate files to be used with your swarm nodes.
-
Upload your Certificate and CA Certificate files to be used with SWCI node used by the SLM-UI.
-
Upload the Machine Learning model file.
-
Upload the SWOP Profile schema file.
Click on the open icon to view the uploaded file.
-
Upload the SWCI tasks file.
Click on the open icon to view the uploaded file.
NOTE:
- For standard examples, this artifacts are available as part of tar file. User can extract the examples on their laptop.
- In a two node example, the network names should be
host-1-net
for the sentinel node andhost-2-net
for the non-sentinel node, respectively.
-
-
Click Save Project to create the Project.
Parent topic:Running Swarm Learning examples using SLM-UI