Update and rename airr_official_1.0_practice_prompt_set_release_publi…

[mikek-mlcommons]

…c_subset.csv to airr_official_1.0_DEMO_prompt_set_release_public_subset.csv

Adding DEMO to name to make it clear this is the demo prompt subset of the larger practice prompt set.

[github-actions]

MLCommons CLA bot All contributors have signed the MLCommons CLA ✍️ ✅

[wpietri]

Unless you think otherwise, say we set this repo up so you are allowed to just merge things.

[rogthefrog]

@mikek-mlcommons @wpietri some of the code in modelgauge depends on the prompt set file names having a specific structure and they are case sensitive. This change should match that structure to prevent potential bugs in case we ever upload that file to modellab and reference it in tests, which has happened before.

I'm not in favor of file names being treated as structured UIDs and manually-managed file names in one system being hardcoded in another system, but it's what we have to work with now, so we should stick to that structure.

Reference:

https://github.com/mlcommons/modelbench/blob/602c62da73298b4ce683fb0c6d26af9d8d10bff1/src/modelgauge/prompt_sets.py#L22

My proposal for this file name:

Instead of:

airr_official_1.0_DEMO_prompt_set_release_public_subset.csv

Use:

airr_official_1.0_demo_prompt_set_release_subset.csv

"demo" in lowercase, and "demo" already implies public.

[mikek-mlcommons]

@wpietri: Thanks!
@rogthefrog All lower case works for me if that's the standard elsewhere. I will rename this to match.