Prior to its closure, USAID was one of the world's largest funders of conservation work — per a recent @insideclimate piece featuring @AidData Chief Economist @ariel_benyishay . 🧵
As the U.S. retreats from #ForeignAid, China is pushing forward, including for women’s development. But where & how has #China supported women’s development abroad thus far? Most projects support #health and job #training, with a focus in African countries. #IWD2025
📊 Happy International #OpenDataDay!
We’re celebrating the power of open data to drive #aidtransparency, accountability & better decision-making
IATI’s #opendata standard helps governments, donors & civil society track development resources & results🧵👇
🏟️Announcing @MistralAI Saba, our first regional language model.
- Mistral Saba is a 24B parameter model trained on meticulously curated datasets from across the Middle East and South Asia.
- Mistral Saba supports Arabic and many Indian-origin languages, and is particularly strong in South Indian-origin languages such as Tamil and Malayalam.
Proceeds from this Wednesday, 02/05, from our DC and MD restaurants will be donated to support @WCKitchen’s humanitarian efforts around the world.
You can also donate directly: https://t.co/4pzGPSIaz4
Amazing what a little bit of data cleaning can do to a loss curve. Fixing about 2% of my labels for a small image classification model resulted in this night and day difference. Can you tell which loss curve is which?
Quick update to my last blog post: I've now included max pooling. It could be effective if some of the dimensions in BERT's embeddings aligned with your classification concept, but this appears to not be the case for gender-responsive aid.
https://t.co/q47Nh9Q5f6
I initially conducted this exploration to satisfy my own curiously while working on NLP models to classify gender-responsive aid, but I thought the results were worth writing up separately.
https://t.co/Jkpdt5OWaa
Give it a read and let me know what you think!
My new blog post is live, describing how to build an NLP regression to quantify climate finance using the World Bank project API. Check it out if you're interested in climate finance or machine learning! https://t.co/cX6LQlihgV
Early voting has started in Maryland! If you need a guide to make informed decisions about the ballot questions in Prince George's county, I recommend https://t.co/PVIvkGrKpA
The #humanitarian funding gap was the largest on record in 2023, and funding is set to fall in 2024. Our new report explores the latest #data on global humanitarian assistance and assesses progress by the #GrandBargain on improving humanitarian action
https://t.co/3TLGw58JRK
DI's Director of Influence and Growth @FionnaCSmyth on how @AnnelieseDodds and the new #UK government can show leadership on #development and rebuild trust with partners in the Global South:
https://t.co/rR09cFdXym @bondngo
THREAD: The @aidtransparency#2024Index results are out. See all the results from the 50 donors assessed.
https://t.co/PoChL9DDXW
Follow the launch event at 9:30 EDT, 2:30 BST today: https://t.co/ypHgAzkez4