Argilla

Today is a huge day for open source AI: Argilla is joining Hugging Face 🤗 🚀 It's time to double down on community, good data for AI, product features, and open collaboration. We're thrilled to continue our path with the wonderful Argilla team and a broader team and vision, with shared values and culture! Thanks to our investors Zetta Venture Partners (James Alcorn), Criteria Venture Tech (Roma Jelinskaite, Albert Morro, Aleix Pérez), Eniac Ventures (Hadley Harris, Dan Jaeck, Monica Lim), and many others, so lucky to have worked with you! https://lnkd.in/dfxvgpsT

Thrilled to see FinePersonas, our latest open dataset, is about to reach 300 ❤️ on Hugging Face If you're interested in synthetic data generation, give it a try! https://lnkd.in/d4rkfD3h

Turns out 40% of the Reflection-v1 dataset had duplicated prompts I’ve shared a deduplicated version and started a discussion here: https://lnkd.in/e5KzPqEQ Let’s invest more time in understanding data and build better open datasets together.

Great to see Argilla datasets and distilabel used and cited in this new RLHF work from AI at Meta! Paper: https://lnkd.in/ej6s7m7Y If you don't know distilabel: https://lnkd.in/dKReQ4_t If you don't know Argilla: https://lnkd.in/dhcaUqip

The formats supported by Argilla are becoming endless. What does this mean? Whether you want to monitor, correct, or enrich your data with new examples to improve your model's performance, Argilla has you covered — no matter the format of your data. Images, documents, videos, audio file, or text conversations—Argilla can process them all. The cherry on top: they're all mixable. I haven’t even mentioned the new "format" feature we're about to release. This will open up infinite possibilities for annotation scenarios! Stay tuned, we’ll release next week. -- 👋🏼 I'm Amélie With my wonderful team, I co-create Argilla—an open-source app for those who want to do AI projects responsibly by taking care of their data. To start with Argilla: 👉 https://lnkd.in/dfYieehh To connect with the community: 👉 https://lnkd.in/eus3YFWW

Why is the FinePersonas dataset 🔥? It can be used to synthesise data of the customer personas your company is interested in! dataset: https://buff.ly/4emlKjC synthetic data library: https://buff.ly/3MXAB8G

Not much of a sharer myself here, but I am delighted to announce that we have just released the Salamandra family of models. 2B- and 7B-parameter multilingual LLMs trained from scratch at Barcelona Supercomputing Center in 35 EU languages code. We are also releasing a first proof of concept of two instructed models (one per size). As the Evaluation & Annotation Lead, I must acknowledge the remarkable work done by Irene Baucells de la Peña, Anna Salles, and Ferran Espuña Bertomeu, who have worked hard to make sure Salamandra models are released with a comprehensive, human-in-the-loop, high-quality evaluation. Many thanks also to the rest of my colleagues at the Language Technologies Unit for their work. We will soon release a technical report with all the details. Next steps include a 40B-parameter Salamandra, the annotation of more curated instruction data, and an aligned model using Argilla as annotation platform. Details and links to the models here: https://lnkd.in/ddhH25Ax

I can't be more excited about the upcoming Argilla release! It will unlock hundreds of new use cases! Stay tuned this week!

How do you start your text classification project on the Hugging Face Hub? David Berenstein will guide you through the journey of creating a text classifier from scratch using Open Source tools. 🚀 Agenda: - Deploy Argilla on Hugging Face Spaces - Configure and create an Argilla dataset - Use model predictions to accelerate labeling - Train a SetFit model 👇🏾Link to the event in the comments

Looking to start a NER project in a optimal way? Nowadays there are good options for zero-shot and more efficient training. zero-shot GliNER: https://buff.ly/3XWO9af efficient training: SpanMarker: https://buff.ly/4eDFDm6 human feedback Argilla: https://buff.ly/3iNecyr