[FEA] Ability to export cuML models to run them on AWS Lambda

[saffranzachary]

Is your feature request related to a problem? Please describe.
I wish I had the ability to launch a cuML model on AWS Lambda. For a mixture of reasons, this isn't possible.

Describe the solution you'd like
I'd love to be able to export cuML models to their Scikit-Learn equivalents, or to ONNX models.

Describe alternatives you've considered
If there were a way to use pip to install a variant/subset of cuML for inference on CPU, that would work just as well.

Additional context
While I think launching cuML models on AWS ECS is a great option, having the flexibility/portability to run a cuML model on more restricted machines would be a very powerful feature.

[github-actions]

This issue has been labeled inactive-90d due to no recent activity in the past 90 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed.

[beckernick]

I believe AWS Lambda doesn't support GPUs, which perhaps makes this request a duplicate of #3556 ?