fairlearn.org - everything related to the Fairlearn package. Please feel free to ask any kinds of questions!
@trivenigandhi @LisaIbanez @brkifle @WesDeng @kevinrobinson during yesterday's community call we discussed some ideas for fairlearn's roadmap. As frequent attendees we'd love to get your opinion as well, so if you have any ideas on where you would like to see the project going, please feel free to reach out! :D
9 replies
(i apologize in advance if i forgot somebody)
Since v0.6.0 is out I moved us from master to main as the default branch. To make things work with
main please approve my PR fairlearn/fairlearn#694 @hildeweerts @adrinjalali @riedgar-ms @MiroDudik @mmadaio
2 replies
Hi all.
With my PR on the correlation removed out there, I feel like I might start working on some educational content for fairlearn. My plan is to host it on https://calmcode.io but I imagine that content could be re-used on the docs as well.
There's one thing though, I don't have a clear intuition on how the reductions actually work. The docs host an example that's a bit too theoretical for me to immediately get my head around ... so is there a call that I can join with a whiteboard soon-ish? I'm not sure if I'm the only one who feels this way, or if the usual meetings are the best place for this, but I'd certainly appreciate a better "paper-napkin"-level of intuition.
4 replies
Hi everyone! In this Thursday's community call, we will discuss @hannawallach 's recent work on fairness through the lens of measurement modeling and think about ways to translate these ideas into a toolkit/educational materials. I think this will be a super interesting discussion, here is a link to the paper: https://arxiv.org/abs/1912.05511?
PR to cover a testing gap I noticed today:
fairlearn/fairlearn#701
We didn't have any purely public tests of the
fairlearn/fairlearn#701
We didn't have any purely public tests of the
MetricFrame with sample parameters
@romanlutz could you please add the link to the notes (https://hackmd.io/QcJ8WBAbQTq6iOgoi0JT4g?both) to the calendar invite?
2 replies
As you may know we have fairlearn.org as the domain name. So far https was not set up for that, so I'm currently trying to do that. If you have issues accessing the page in the meanwhile please respond by creating a thread here. https still doesn't work since GitHub first needs to issue a certificate, but http://fairlearn.org and fairlearn.github.io should both work and redirect to fairlearn.org (watch what your browser resolves to). Works at least on my computer and my phone. Https should work later tonight/early tomorrow depending on which timezone you're in.
6 replies
I'm trying to figure out what the vibe around the
model-card-toolkit is. Anybody here knows anything? Seems like there isn't really a community around it. I'm wondering if we have any contacts in the team that we could talk to, to see what their plans are. Also wondering if it makes sense or if there's enough support to actually fork it and create a community around it.
12 replies
It also has a hard dependency on a bunch of tensorflow related packages, which I don't think is necessary for a model card toolkit kinda library.
Thanks everyone for the interesting discussion! I felt like I needed a brain dump, so I opened a discussion (https://github.com/fairlearn/fairlearn/discussions/707)
Please feel free to add your thoughts, perhaps its helpful to collect some of the things we're thinking about to structure next week's followup :)
Please feel free to add your thoughts, perhaps its helpful to collect some of the things we're thinking about to structure next week's followup :)
Hi! A friend of mine is working on a study on expressions of emotions in open source communities. He is currently looking for people who regularly contribute to OS for an interview study. In case any of you are interested to participate, here's a link: https://www.linkedin.com/posts/nathan-cassee-a92a9b154_interview-on-open-source-contributions-activity-6772461268994469889-HVc5
4 replies
Following yesterday's call I put some of the teaching materials of our responsible data science course on github. I don't feel super comfortable yet putting them on a public repo (blabla perfectionist etc.). If anybody would like to have access just respond to this thread or send me a DM. There's a few tutorials in there and lecture notes with an intro on algorithmic fairness.
We All Count (a super interesting project that tries to help people with increasing equity in data/statistics projects) has something called "talk to your boss sheets": https://weallcount.com/talk-to-your-boss-sheets/?mc_cid=dd0c1015c7&mc_eid=3ec8760ca5
5 replies
Above is the report! I believe Dr. Haiyi Zhu already shared it with @romanlutz and @MiroDudik !
I seem to have received 2 invites, and I'm currently waiting in the MS Teams app.
Any objections against adding this? https://github.com/apps/goodfirstissue somebody pointed this out on Twitter. I'd also like to see if we can curate a few good first issues. We have a bunch tagged as
@adrinjalali @hildeweerts @MiroDudik @riedgar-ms @mmadaio
help wanted but not all of them are good first issues, and we probably have many potential tasks that aren't listed as issues at all. If anyone has thoughts lmk!@adrinjalali @hildeweerts @MiroDudik @riedgar-ms @mmadaio
4 replies
@adrinjalali I'm wondering how the model cards relate to the visualizations from Kevin's PR that I took over. Obviously I've been slow to address the feedback, but if you manage to get the model cards to work is there still a need for anything else? I'd prefer extending the model cards then rather than add several kinds of visualizations. Of course, I'm well aware that model cards are about much more than just visualizations, but I'm restricting myself to that aspect right now since it relates to the visualizations PR.
2 replies
Any idea why this links to Fairlearn? https://paperswithcode.com/paper/fair-bayesian-optimization
3 replies
The current iteration of the webpage is here: https://2469-133444044-gh.circle-artifacts.com/0/docs/_build/html/index.html
The use case text is to be slightly updated following https://github.com/fairlearn/fairlearn/pull/709#pullrequestreview-616285274
If you have any other suggestions how to improve the wording, please comment on the PR!
@LisaIbanez : When I tried to pare down the "fairness is sociotechnical section" the visual balance between the left and right column was off, so I'm in favor of keeping it for now. Is there some other useful information that we could add there instead?
8 replies
Hi everyone. I found fairlearn after watching a couple of talks on algorithmic fairness and I'm interested in contributing and learning more about the field. I was going to ask about if fairlearn/fairlearn#254 was still a good first issue (it's been a year since it was updated and the docs seem to be updated with return types from my cursory glance). If it's not, are there any other 'good first issues' that I can get started on that maybe haven't been updated on the issues section yet?
5 replies
Would it make sense to add a separate label for github issues related to the user guide? I currently put these all under documentation but i'm not sure whether that makes sense. Same for issues related to the website (content/design)? I'm probably overthinking this :D
6 replies
I've got a happy announcement. It seems like FreeCodeCamp will be using my instructional material on https://calmcode.io to teach scikit-learn. What's great about this is that these videos contain an example with
load_boston with a full acknowledgement of the issues. It will be the first, to my knowledge, training material to do just that.
2 replies
Behold! A scikit-learn tutorial that acknowledges the load_boston issue. Available now ... on freecodecamp.
https://www.youtube.com/watch?v=0B5eIE_1vpU&ab_channel=freeCodeCamp.org
Please can I have some reviews for:
fairlearn/fairlearn#727
I'd also like to do a bugfix release (it would be v0.6.1) of Fairlearn for this.
fairlearn/fairlearn#727
I'd also like to do a bugfix release (it would be v0.6.1) of Fairlearn for this.
9 replies
And a PR to fix the PR-Gate. Looks like the path to
fairlearn/fairlearn#735
yarn changed on the build machines:fairlearn/fairlearn#735
Fairlearn v0.6.1 is out:
https://pypi.org/project/fairlearn/
This is a bugfix release, mainly for the integration point with AzureML (special case code, which will actually be removed soon)
https://pypi.org/project/fairlearn/
This is a bugfix release, mainly for the integration point with AzureML (special case code, which will actually be removed soon)
There's one considering adding
count which may be a good start for you