climate-tech-handbook / data-magic

Scripts and other fun tricks to enhance the site
MIT License
10 stars 5 forks source link

Figure out dataset training strategy #17

Closed futuresoup closed 1 year ago

futuresoup commented 1 year ago

We need to reference the data in #19 and build a pipeline to gather and train our AI.

@Yinan0409 has made progress here.

How do we continuously pull all the resources listed in #19 and generate the appropriate csv files through automation?


Another approach worth exploring is to use Llama Index for training inside of our Flask App.