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

ML model handling #13

Open
jetschny opened this issue May 15, 2023 · 2 comments
Open

ML model handling #13

jetschny opened this issue May 15, 2023 · 2 comments
Assignees

Comments

@jetschny
Copy link
Contributor

as for now it seems that the resource meta-data is target to document which ML algorithm has been applied to create which data but how do we handle the documentation of the ML model as such? can we have a review of how do document a ML model? we would need to have a switch field somewhere if it is a model or data as output and we then we could use most of the fields but instead of a data name we give a ML model name?

@KathiSchleidt
Copy link
Member

KathiSchleidt commented May 16, 2023

@jetschny Trying to understand what you're actually asking. I'd initially thought it's details on describing an ML Model, but the statement on we would need to have a switch field somewhere if it is a model or data as output... confuses me.

I do see quite a bit missing on model configuration in the current resource metadata

@sMorrone sMorrone assigned cozzolinoac11 and unassigned sMorrone May 16, 2023
@sMorrone
Copy link
Contributor

@jetschny the idea underpinning current resource metadata structure is that a ML algorithm must be evaluated/documented in the context of the specific case where it is applied. The "reference link" field is meant to provide link(s) to available resource documentation e.g., a web page describing the ML algorithm (in the issue #6, for example, the reference link is https://en.wikipedia.org/wiki/LeNet and describes the LeNet Classifier DL resource)
Is this a reply to your question, or did I get you wrong?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants