Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Update progress bar for PAR fitting (+ save loss values) #80

Closed
npatki opened this issue Aug 30, 2023 · 0 comments · Fixed by #82
Closed

Update progress bar for PAR fitting (+ save loss values) #80

npatki opened this issue Aug 30, 2023 · 0 comments · Fixed by #82
Assignees
Labels
feature request Request for a new feature
Milestone

Comments

@npatki
Copy link

npatki commented Aug 30, 2023

Description

Currently, the PAR model's verbose=True option does a good job at showing me the progress bar and the loss values. However, it is not streamlined with our ideal usage and it also isn't possible to get the loss values later.

Expected behavior

Progress bar: When verbose=True, continue to show the progress bar as usual, but with some minor modifications.

  1. There is no need to show the Epoch #, as that is what the progress bar is incrementing by
  2. Round the loss values to 3 decimal digits, in order to keep to consistent layout
  3. Instead of showing the time per iteration/elapsed time, show the estimated time remaining instead

Save the loss values for later use: The full (unrounded) loss values should be saved for future use. They can be accessed through a parameter. Save them as a pandas.DataFrame.

>>> par.loss_values
      Epoch     Loss
0     1         -1.0191212
1     2         -1.23012421
...

Additional context

See CTGAN issue sdv-dev/SDV#298

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
feature request Request for a new feature
Projects
None yet
Development

Successfully merging a pull request may close this issue.

3 participants