Jupyter Notebook documenting the use of earth-text

[lauracchen]

Just asked @yellowcap but figured I'd put it here -

Just got to see the video summary of where you're at. Thanks so much for that! To get started with coordinating with the app team on our end -

@alkalait @rramosp
Could you all create a demo widget that takes tags as input and allows you to look at the images associated with the embeddings you get back through similarity search? Ideally this also shows how many instances of the tag that occur in the image, and the area covered by the tagged polygon. The app team is hoping to get a sense of how this is performing, and be able to play around with it

cc @lukaszknasiecki

[yellowcap]

I think demo widget would mean Jupyter notebook in this case (as agreed on our call last week).

[yellowcap]

I updated the title here to be more explicit of what is to be done.

[lauracchen]

Thanks - can you confirm what this notebook will cover? Basically selection of tags as input and going through steps to ultimately produce associated chips/images? Will you also be able to output the associated info of the surfaced chips, i.e. counts of tags, areas, perimeters?

@alkalait @rramosp @yellowcap can you also let us know your estimated landing date for this? Trying to coordinate with what's possible on our app side. Thank you!!

[alkalait]

Hi @lauracchen

The following notebook

https://github.com/Clay-foundation/earth-text/blob/rramosp/notebooks/models/11%20-%20full%20query%20workflow.ipynb

demonstrates the full autocomplete and search workflow as explained in our previous calls.

Please let us know if you have any questions.