search

containers / podman-desktop-extension-ai-lab

Work with LLMs on a local environment using containers
https://podman-desktop.io/extensions/ai-lab
Apache License 2.0
182 stars 40 forks source link

Implement E2E test scenario to cover the creating of a model service #2100

Open odockal opened 1 week ago

odockal commented 1 week ago

Is your enhancement related to a problem? Please describe

In this scenario we want to verify the user e2e path of model service. This is a smoke e2e test case, where we will test basic path for the model service creation and application check.

Describe the solution you'd like

Implement a e2e test using playwright covering:

  1. AI-Lab is installed, podman machine prepared (win, mac)
  2. Open Catalog Page
  3. Download a model from catalog (ggerganov/whisper.cpp < 0,5GB) - find in the table, click on download, wait for model to be downloaded, check it is present
  4. Open Services -> New Model Service
  5. on Creating Mode Service Page -> Choose model from dropdown, press Create Service
  6. Verify all steps are performed (inference server is up, model copied, image pulled, etc.)
  7. Open Service Details
  8. Check all label (endpoint url, CPU/GPU inference, model, client code)
  9. Send request to a running service (curl https://localhost:44331) -> returns "Hello" with html.
  10. Stop service
  11. Delete service

Describe alternatives you've considered

No response

Additional context

Many issues depends on platform we will run tests on. On windows we might need to switch uploading of a model to a machine.

Also, create issues to cover other paths (restarting service, creating servince from Model's Catalog, etc.).