The developer community is currently showing significant excitement towards DSPy, a new framework developed by Stanford NLP researchers (@lateinteraction). DSPy is designed to facilitate the building of LLM-based applications, marking a shift from merely prompting language models to programming them. This advancement is seen as a game-changer, with the potential to focus future announcements on entire systems rather than just new LLMs, including developments like Perplexity and ChatGPT. DSPy allows for the compilation of new language models, showcasing its capabilities through applications like Gemma. The technology stack involved includes notable names such as Google, Hugging Face, and Weaviate. A session to introduce DSPy and its application in building a simple RAG system with Clarifai is scheduled for 10:30 am ET, highlighting the framework's practical utility.
Really cool example of a compound AI *system* that enables new AI capabilities, and it’s built using DSPy. I expect we’ll see a lot more like it this year. https://t.co/mh4NokxVhj
Compiling new language models using DSPy is one of its superpowers. Awesome to see @helloiamleonie applying this to Gemma! Btw look at that tech roster (Google + Hugging Face + Weaviate + DSPy) 😮💨 https://t.co/UlM3LheBfK
Compiling new language models using DSPy is one of its superpowers. Awesome to see @helloiamleonie applying this to Gemma! btw look at that tech roster (Google + Hugging Face + Weaviate + DSPy) 😮💨 https://t.co/UlM3LheBfK