Numinous deployment #19079 fails with MODEL_RUNNER_CONNECTION_FAILED

Hi CrunchDAO team,

I am participating in Numinous: Predictive Agents For Real World Outcome. The notebook submission and deployment were created, but the deployment failed at the runner connection step.

Project:

Model:
general-thrush

Submission / deployment:

  • Submission #1
  • Deployment #19079
  • Model ID #15612
  • Runtime type: CPU

What appears to work:

  • The notebook upload/conversion completed and produced main.py, notebook.ipynb, and requirements.txt.
  • Builder logs reached POST_BUILD / artifact upload success.
  • Runner logs include: ModelRunner started and ready to serve on [::]:50051 in mode secure.
  • Local artifact validation and local replay checks pass.

The deployment page still shows:

MODEL_RUNNER_CONNECTION_FAILED: Failed to establish a connection to the model runner. We are informed of the issue and will resolve it as soon as possible.

The deployment remains in a runner failed state, with expected status stopped.

Could you please check whether this is a platform-side runner connectivity issue for Numinous deployment #19079, and whether I should wait for recovery or redeploy after a fix?

Thanks.

You are not the only one to receive this error.

Just restarting or re-submitting would be enough.

Network drop sometime happen.