Skip to content

Evals as part of CI #2288

@leshy

Description

@leshy

It can easily happen that someone PRs a core change with deep performance consequences downstream

we have special tests that profile execution speed, CPU and MEM footprints of certain components (mapper, maybe detector, sometimes full blueprints, startup speed), how/where could we log this data? as well as fail tests upon significant deterioration?

what is the standard output for benchmarsk like this? basically a number per key.. lower is better, how do we format this?

would be nice also to get historical graphs for benchmarks to see the repo health, even individual users contributions

Metadata

Metadata

Assignees

Labels

No fields configured for issues without a type.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions