Using the grants data lake?

[thelastjosh]

Hey @ryscheng @ccerv1 , given the existence of DAOstar/OSO grants data lake, I'd like to understand:

• is that grants data lake useful for this contest?

IF IT IS USEFUL, THEN:

• We should advise the people putting out this contest (if that's not yourselves) about the existence of the grants data lake.
• We could define a bounty to support additional work on the DAOstar/OSO grants data lake that makes it better for this contest. In particular: $2k USD to (1) import the underlying contest data sets (OP retro, Open Collective, etc.) set into the existing data lake, (2) export the contents of the data lake into the competition format, i.e. an extended competition data set, (3) all notebooks / scripts used to produce the above. Doing this should have obvious utility for expanding the impact of the contest.
• We should submit the entire data lake to the contest.

FYI Rashmi, Josh, and I were just discussing this earlier in the day.

Also, if I don't catch up with you in the next few days: happy holidays!!

[ryscheng]

100% useful! Quick summary of where we're at:

• Existing grants are the best source of training data for deep funding.
• All of the training data sets that have been produced thus far originate from our data lake! This repo is chock full of notebooks that mostly Carl's been producing for deep funding. For example, check out this notebook

So yes, 100%, we should try as best as possible to advertise the data lake as a rich source of data for ML training.

We should definitely collaborate on improving the data lake for the contest, since we're the primary source of data at the moment! Will respond in TG with more thoughts in a bit