Skip to content

Model Pipeline #23

Description

@KurbyDoo

Background

The model pipeline describes our main entrypoint for creating clean patches from noise. Callers send in a noisy patch of fixed size, then get back a cleaned patch of the same size. Here, we load the models from #16 and pass inputs between the two models. This module will be called by the #4 for every region it cuts.

Description

To have a barebones workflow we should be able to give a patch and return a patch of the same size

  • Load mock models from inference
  • Connect model inputs and outputs to make a full resolution output
  • Apply the elevation encoding and return the output

Metadata

Metadata

Assignees

Labels

No labels
No labels

Projects

No projects

Relationships

None yet

Development

No branches or pull requests

Issue actions