Sakari Ikonen

regarding possible changes to `sanitize_for_argo` and its surjectivity in order to not have collisions between deployed flow names, this seems somewhat of a tricky issue. 1. if we stick to...

As a heads up I'll be looking into some other inconsistencies of `@resources` as well next week that have been on the backlog for a while now, and will hold...

> @saikonen, I just wanted to check in on this PR. Is it still of interest, or should I go ahead and close it? I removed all the import-sorting fixes...

note that for testing this, both `.pip/metadata` and `conda.manifest` local files for the involved flow need to most likely be cleared

I think the existing print statements might be leftovers from way back. The preferred way to fix this issue should probably be to utilize the preconfigured logger from `services.utils` wherever...

noting something extra that needs to be handled here: ```python from metaflow import step, FlowSpec, resources import os class S3WorkerFlow(FlowSpec): @resources(cpu=1) @step def start(self): val = int(os.environ.get("METAFLOW_S3_WORKER_COUNT")) assert val ==...

> running this `--with kubernetes` or argo workflows fails, as resource decorator values are cast to a str-float, which fails to convert back to an int The cause for this...

tested this with `@parallel` and seems to be working as expected so we can cross that off the list.

I'm going back and forth regarding the UX with `spec:` On one hand exposing the controls seems like a valid use case for fine-tuning the ephemeral storage. On the other...

The issue is relevant and the PR solving this is going in the right direction. I'm currently looking into why the default gpu value for resources is `"0"` to see...