Spaces:

Langflow

/

Langflow

like 948

Running

1. The character splitter could accept a regex, giving the splitter an enormous amount of robustness. Might have to change some implementation to use the regex to it's most potential... but still just allowing a regex would make it like a character splitter/parser
2. Complete the placeholder text for all the fields that are not obvious, and eventually all fields. Some fields are not so obvious what is asked and/or what format.
3. Use dropdowns where appropriate eg. OpenAI Model

I've never changed anyones code on GitHub so that's why I'm not doing it myself... don't want to make a mistake