I'm really excited about this new offering from DataKind's Center of Excellence! Vote in the poll or comment below to let us know what topics we should be covering to best support social sector organizations...

Join us September 16 for our 10 Year Anniversary virtual event, celebrating the Power of a Decade of #dataforgood, #DataKind's impact using #datascience and #AI to advance the social sector, and exploring what the future can bring together. bit.ly/3Q1MyJr #DKx10 #UNGA77

We're thrilled to launch #Nonprofit #Data 101, a free webinar series + office hours for social sector professionals! Our first webinar will be on Wednesday, August 24, 10am ET, covering Data Project Scoping. For more info. and to register: bit.ly/3vwcPYn #DataKind

Join us September 16 for our 10 Year Anniversary virtual event! We'll celebrate the Power of a Decade of #dataforgood, #DataKind's impact using #datascience & #AI to advance the social sector, and explore what the future can bring together. bit.ly/3Q1MyJr #DKx10 #UNGA77

!!!! Ok I recorded a (new!) 2h25m lecture on "The spelled-out intro to neural networks and backpropagation: building micrograd" youtube.com/watch?v=VMj-3S… . This is the culmination of about 8 years of obsessing about the best way to explain neural nets and backprop.

Introducing "This JACS does not exist": I trained a bunch of AIs to generate chemistry paper ToCs with relevant titles and abstracts! Check it out here: thisjacsdoesnotexist.com and twitter bot This JACS does not exist.

📈 New (1h15m) video lecture (#3): The spelled-out intro to language modeling: building makemore. Part 2: MLP youtube.com/watch?v=TCH_1B… > We continue our implementation of makemore: the multi layer perceptron (MLP) language model of Bengio et al. 2003

Join us for our first in person DataDive in over two and a half years at Georgetown University on October 22! meetup.com/datakind-dc/ev…

Closing out our Data Dive!

Our first social event since the pandemic!!!!

Been working on LLMs in production lately. Here is an initial thoughtdump on LLMOps trends I’ve observed, compared/contrasted with their MLOps counterparts (no, this thread was not written by chat gpt)

At a talk at Data Viz DC by Juan Vega about work done in partnership with DataKind DC and the nonprofit starting with today

Take a look back at DataKind DC's first-ever Citizen Science #DataDive where more than 50 volunteer #datascientists worked with National Geographic, Audubon Society, Earth Challenge 2020, among others. Check it out! bit.ly/2XbvzJK #WorldOceansDay #dataforgood

Off to the start of a Data dive at Georgetown partnering with the Data Science and Analytics program!

At the capstone project for Georgetown's DSAN capstone projects. Looking forward to seeing their work for Hildreth Institute

At the start of another DataDive at Georgetown!