1. 20
  1.  

    1. 2

      The shorthand syntax is comfortable. I may end up using llm schemas dsl all by itself—thanks for including it. For local models and those whose services don’t support a schema, I imagine the user could construct a suitable prompt template with that ingredient.

      1. 1

        As someone who spends a lot of time wrangling with scanned documents and OCR output, I’m excited to give this a spin.

        1. 1

          In my experience, the most useful LLM feature. Now able to submitted an hypothesis about relationship between a malady (disease; allergy; drug:drug interaction) and factor inputs (food; drug; supplement) and receive a response in the same format as used by a datastore (JSON schema). Before structured data support, an LLM response had to parsed and reorganized to combine with existing data structures.