Replies: 2 comments
-
I think what we want is to document the code really well and write the code in AI friendly manner such that a code assistant running in the editor can easily figure out what's needed and/or what to do |
Beta Was this translation helpful? Give feedback.
-
I agree with that. The detailed docstrings help no matter where you use the AI. My thinking of the GPTs, was someway to get it more interactive api/how-to-reference, for the pure users that don't download the git repo. And only know a little python, then we use the GPT to trick them in to learning python and contributing to the repo :) Basically a slower user learning curve, that doesn't cost us anything. Also, the GPTs are manual upload/edits only, so scratch that automation. |
Beta Was this translation helpful? Give feedback.
-
I made a gpt assistant for our api docs.
Here is the manual upload version: https://chatgpt.com/g/g-67b9dffbaa988191a4c7adfd4f96af65-particula-assistant
@mahf708
Right now, it is good for api references. If I add our jupyter notebooks, it might help in writing new simulations.
End stage vision would be automatically update the reference material in an open ai, gpt?Say on each PR that gets merged?GPTs are manual only.
Beta Was this translation helpful? Give feedback.
All reactions