Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Suggestion] Loading torch-model-archiver arguments from YAML #780

Open
skat00sh opened this issue Nov 10, 2020 · 6 comments
Open

[Suggestion] Loading torch-model-archiver arguments from YAML #780

skat00sh opened this issue Nov 10, 2020 · 6 comments
Assignees
Labels
enhancement New feature or request good first issue Good for newcomers triaged_wait Waiting for the Reporter's resp

Comments

@skat00sh
Copy link

Is your feature request related to a problem? Please describe.

torch-model-archiver takes in a lot of params that makes it a hefty command line argument

Describe the solution

Allowing all the param values to be parsed from a YAML file when torch-model-archiver is used with a --yaml flag.
It will also help automating creation of multiple .mar from different configurations.

Describe alternatives solution

I've tried calling torch-model-archiver from a shell script with variable params and that works file as well. But I just wanted to bounce this idea off the community to see if it is something people could find useful.
I'd be happy to work on it.

@skat00sh
Copy link
Author

@dhaniram-kshirsagar @harshbafna Any thoughts on this?

@dhaniram-kshirsagar
Copy link
Contributor

@skat00sh Yes, it will be nice to have YAML/json as an alternative to command input.

@skat00sh
Copy link
Author

Thanks @dhaniram-kshirsagar! Will start working on it soon

@msaroufim msaroufim added enhancement New feature or request good first issue Good for newcomers labels Apr 22, 2021
@msaroufim
Copy link
Member

Hi @skat00sh let me know if you're still interested in working on this so we get something merged. This involves 4 steps

  1. Python has a nice YAML library https://stackabuse.com/reading-and-writing-yaml-to-a-file-in-python/
  2. And would need to update the arg-parser https://github.com/pytorch/serve/blob/master/model-archiver/model_archiver/arg_parser.py
  3. Adding a sample YAML config in this directory https://github.com/pytorch/serve/tree/master/model-archiver
  4. Testing the model with a YAML config and validating that it runs

@msaroufim msaroufim added the triaged_wait Waiting for the Reporter's resp label Apr 22, 2021
@skat00sh
Copy link
Author

@msaroufim Sorry, I almost forgot about this issue! I'm back on it now, will finish it asap! Thanks for defining the steps!

@skat00sh
Copy link
Author

@msaroufim I think I should release it for now for someone else. With the academic workload this sem, I'm finding it a bit hard to find time for anything. Sorry that I took so long again to revert.

I wished to mention 2 more things though:

  1. Is there a chat-group or discord server like thing that I could join to be a part of the development in the near future? I find the revert latency at github issues sometimes is a bit too much (I'm guilty of it myself)
  2. While trying to setup the dev environment, I faced a couple of issues. I unfortunately didn't log them but if I'm working on it again, what do you think should be the quickest way to reach out to devs?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request good first issue Good for newcomers triaged_wait Waiting for the Reporter's resp
Projects
None yet
Development

No branches or pull requests

3 participants