Adala

We recently open sourced an agent framework [1] for automating data processing and labeling where the agent's prompt is refined trough iterations with the environment and then asking to an LLM to revise the prompt according to its performance (i.e. automatic prompt tuning). We tested it on the Math reasoning dataset GSM8k and where able to improve the baseline accuracy (GPT4) by 45% -> 74% using 25 labeled examples (I'll put the notebook and blog post linked below [2][3]). Results are definitively very interesting, if not surprising with some skills, and we see more and more of our open source users and customers showing interested in the framework for automating labeling / having it as a copilot. [1] https://github.com/HumanSignal/Adala [2] https://github.com/HumanSignal/Adala/blob/master/examples/gs... [3] https://labelstud.io/blog/mas

Hello HN! We are thrilled to introduce a project we've been working on for quite some time – Adala, an open source *A*utonomous *DA*ta (*L*abeling) *A*gent framework. Adala is a robust framework for implementing agents that specializes in advanced data processing tasks. Think of these agents as smart assistants that can teach themselves different tasks or skills over time, such as data classification, summarization, or data generation. The agent learns based on the context you provide and the data it encounters. For Adala, you define this context by providing it with a ground truth dataset. True progress in the field of AI most often comes from accessible knowledge, collaboration, and strong feedback loops. By open-sourcing Adala, our goal is to inspire creativity and the development of novel applications. And we want to help drive new st