search

BoxcarsAI / boxcars

Building applications with composability using Boxcars with LLM's. Inspired by LangChain.
MIT License
434 stars 39 forks source link

Do not use the engine_prefix to start the LLM prompt for the XML Train #106

Closed francis closed 1 year ago

francis commented 1 year ago

Before, we would always start a prompt with the engine prefix list:

This would then break up the XML across two records of the conversation. This code keeps the XML clean in one record of the conversation.