Skip to content

Pivot to BayesChime #568

@BrianThomasRoss

Description

@BrianThomasRoss

With the pivot to the new model there are a few points to address

We can publish the original model as a PyPI package - As pointed out by the @ChrisPappalardo
from Django-Chime in the slack - this would let him and likely others do more with the model.

One thing to consider is use in Colab. This pertains to both old and new models alike. Currently the project requires Python >= 3.7, however it's my understanding that the Python version on Colab is 3.6.9.

Another is what do we want in terms of scope on in the PyPI package? The three options I see are:

  • Including streamlit and and the cli and installing everything below that as a dependency
  • The above option except including the entire contents of penn_chime
  • The above option sans the streamlit app

Also to publish to PyPI we require two accounts:

Also clarification regarding meta. Here is the documentation regarding python packaging metadata showing what is optional and what is required.

Thanks for your time

Metadata

Metadata

Assignees

Labels

questionFurther information is requested

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions