RT @motherduck: Leveraging LLMs in data transformation pipelines is now easier than ever with the prompt() function in MotherDuck. https:/…

@david_govea @motherduck @AnthropicAI We also have an eval space on HF to experiment with different prompts and models

Works also with local DuckDB now!🦆🚀

@alexalbert__ ;)

@0xperp @motherduck @AnthropicAI Sure, feel free to DM!

Really wish there was a way to pass more than text between tools. Like an arrow table for example 🤔

RT @krish_adi_: hey @llmsdottxt peeps would be cool to make it open source, so people can add to it.

@calebfahlgren @duckdb Oh that's awesome! I added openai and potion-8M embeddings here, computed with DuckDB of course: Should be fairly easy to wrap that up in a pipeline that runs every now and then?

Huge thanks to @calebfahlgren for the idea and collaboration on the space 🙏 & @NumbersStnAI for their support on the benchmark suite! 🙏

RT @MadelonHulsebos: Interested in making representation learning and generative models work for structured data (e.g. tables in relational…

@RandVegan @duckdb @motherduck It's a custom SQL aggregation function, computation is done within the database engine.

Same data set, local DuckDB vs sklearn, 20x difference - the duck is faast 🦆💨

@JPedro_Monteiro @matsonj Oh yeahh :)