ml-infrastructure

I am using metaflow locally but with the AWS service (e.g. the actual compute is happening locally rather than in AWS batch but the metadata is using AWS). When I access the run data through run.data I get new local directories with names like metaflow.s3.w3efey1k, which I presume is because metaflow pulls from S3 into that directory, and then un-pickles the result from there. Is there a way t

While reading the code I found the following (minor) issue.

The aiohttp docs states: app.make_handler() is deprecated and will be removed in future aiohttp versions. Please use Application runners ins

Add overview and map proper issues to it.

Is your feature request related to a problem? Please describe.
Since there's not much logic, testing needs to be done with integration tests (there isn't much opportunity for unit testing), mocking GCP services and k8s. Especially to check how the failure cases are handled.

Describe the solution you'd like
Given a config for the edge call, we verify the calls that it makes to APIs suc