deprecated in favor of MLM extension

[fmigneault]

The Machine Learning Model (MLM) extension (https://github.com/crim-ca/mlm-extension) combines the fields that were previously defined in the Deep Learning Model (DLM) extension as well as most (all?) fields proposed by ML-Model. Some fields are renamed to avoid redundant details between the 2 references, while others are adjusted to allow more flexibility (e.g.: not just docker-compose runtime, but virtually anything). More best-practices and examples are provided to demonstrate the use of MLM along other STAC extensions to take advantage of the full STAC ecosystem.

Schema for MLM: https://crim-ca.github.io/mlm-extension/v1.0.0/schema.json

Fixes #13

[m-mohr]

Waiting for https://github.com/orgs/stac-utils/discussions/4 to be discussed.

If we go forward with this PR, we should not forget to archive this extension via the GH settings.

[HamedAlemo]

Echoing @m-mohr, based on the outcome of https://github.com/orgs/stac-utils/discussions/4, this should be deprecated.

[emmanuelmathot]

Instead, I think this extension should not be deprecated until the new one increased its maturity level to Candidate.

[fmigneault]

"Candidate" is a long way off. MLM is already at "Pilot", which is 1 level higher than ML-Model has been on for ~3 years.

I need stac-extensions/stac-extensions.github.io#48 to be integrated to reflect the latest changes and increase adoption.

[emmanuelmathot]

Actually, ml-model should be in "Candidate" level but was never reviewed.
Anyway, my view here is to not deprecate so quickly an extension by another which in not actually covering the same purpose.

[fabricebrito]

I'm in favour of keeping this extension as its purpose is valid and needed. This extension doesn't prevent the MLM from existing.
Looking forward for #17

[fmigneault]

@fabricebrito

That's the whole issue and reason about deprecating ML-model.
ML-model is not planned to receive any maintenance, so #17 will most probably not be included.

Instead, efforts should be done in MLM to include any necessary adjustments (if they are actually required) to support all use cases with a single extension. Looking at the PR, I think this is already possible with MLM in its current state, since all ml-model roles have a corresponding mlm role (https://github.com/crim-ca/mlm-extension?tab=readme-ov-file#mlm-asset-roles), and it doesn't impose any limitation about docker-compose for its runtime.

[rbavery]

Hi everyone, Francis and I (maintainers of the MLM) have agreed that the stac-extensions repo is the best home for the MLM extension.

We got approval from CRIM, who owns the current MLM github repo, to move the extension. Given that we don't want to have two competing specs and the MLM can accommodate docker-compose I think this extension should be deprecated.

The MLM extension has evolved quite a bit to meet a range of needs when describing models today. I think we should focus efforts on the MLM once a move is approved to the stac-extensions repo (working on this now in the Gitter channel). We are open to others providing input as contributors and maintainers as we continue to improve the schema and tooling around the MLM spec and would like the MLM to be collaborative and community driven.

[fmigneault]

@m-mohr

This is now my next step in the migration plan:
https://github.com/orgs/stac-utils/discussions/4#discussioncomment-10767685

The repository and schemas are already migrated:
https://github.com/stac-extensions/mlm
https://stac-extensions.github.io/mlm/

Once this PR is merged, I propose also updating the repository description to redirect to MLM, as following:

An Item and Collection extension to describe machine learning (ML) models that operate on Earth observation data (deprecated, see https://github.com/stac-extensions/mlm)

This will update the description information in the https://stac-extensions.github.io/ page to better guide users.