Model serving failing for select models

Incident Report for Baseten

Resolved

Model serving requests for some models running on A10G and T4 GPU types failed with 499 or 503 status codes from approximately 14:50 UTC until 16:05 UTC.
Posted May 13, 2024 - 09:05 PDT

Investigating

This issue has been resolved
Posted May 13, 2024 - 07:50 PDT
This incident affected: Model Inference.