About the study by Devin
About the study by Devin
We like to keep our experiments as simple as possible.
So it's not a good idea to mix multiple stories in monorepo
Don't tie it to a repository, just give them the information they need and let them implement it.
Push the result to a new repo (it's too much trouble to receive it in patch because there is both data and code).
Isn't this enough? it turned out to be
In other words, a REPO is created for each experiment.
Even experiments that are not useful should be recorded.
We don't want experiments that aren't useful mixed in with the next one.
Write details in README
Have Devin write the README itself, but have a human fix it.
When I think about it, it doesn't feel right to do monorepo in broadlistening-research or to put it in kouchou-ai.
relevance
---
This page is auto-translated from /nishio/Devinによる研究について using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I'm very happy to spread my thought to non-Japanese readers.