Artifact Under Review
12.4. Performance Requirements
Team Number for Team Doing the Review
27
Description of Issue
When the model crash, or return an error, what should be our fallback behaviour? Since achieving 99% uptime is extremely hard.
For example, AWS can only ensure the uptime of the containers, but not when there's errors within the code (run-time errors). Our method is to host an open-source model in conjunction with the self-made one, and when our models returns an error, we will query this open-source model.
Artifact Under Review 12.4. Performance Requirements
Team Number for Team Doing the Review 27
Description of Issue When the model crash, or return an error, what should be our fallback behaviour? Since achieving 99% uptime is extremely hard. For example, AWS can only ensure the uptime of the containers, but not when there's errors within the code (run-time errors). Our method is to host an open-source model in conjunction with the self-made one, and when our models returns an error, we will query this open-source model.