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.

Sign up for GitHub

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

Jump to bottom

Plot control #42

Merged
merged 16 commits into from
Oct 24, 2024
Merged

Plot control #42

merged 16 commits into from
Oct 24, 2024

Conversation

alex404
Copy link
Contributor

@alex404 alex404 commented Oct 24, 2024 •
edited
Loading

Added the ability to switch off (very heavy) per channel analysis, and resolved a potential bug with image normalization. Updated default configs accordingly.

Resolves #27 #40

alex404 and others added 13 commits October 21, 2024 17:08
wandb dir now changed to be in the run_dir of the run.
2213ed9
Fixed up directory stuff to allow reloading (in theory)
ebf964c
casting dict_conf for fewer errors
ea236cd
Merge remote-tracking branch 'origin/master' into code-cleanup
cae8119
@alex404
Removed imageset length warning. I seem to be using len safely. Renamed
97b8129 
ImageSet.__len__ to ImageSet.epoch_len to better reflect the function
@alex404
Results of torchinfo now saved to disk, and should upload to wandb as
193a4e0 
well
@alex404
AdamW now default
069bee7
@alex404
Added __len__ back into ImageSet so that it is a "map-style" dataset
76c2665 
(https://pytorch.org/docs/stable/data.html#map-style-datasets)
@alex404
Better arguments for wand.save
97eed06
@alex404
Better normalization of plots
f5df460
Linting error
bc35066
Refactored cnn_statistics, and trying to add a switch to disable
34652fa 
per-channel analysis.
Resolved bugs with channel analysis control. Improved default configs.
cd4134a
@alex404 alex404 requested a review from fabioseel October 24, 2024 06:46
fabioseel
fabioseel requested changes Oct 24, 2024
View reviewed changes
Copy link
Contributor

@fabioseel fabioseel left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

  • I think the config naming of 'simulation' is misleading.
  • in my opinion reconstruct_images should not return the normalization_stats
  • can you check whether the path is correct in runner/analyze.py:89 ?

resources/config_templates/user/experiment/classification.yaml Outdated Show resolved Hide resolved
retinal_rl/analysis/statistics.py Outdated Show resolved Hide resolved
runner/analyze.py Outdated Show resolved Hide resolved
runner/analyze.py Outdated Show resolved Hide resolved
runner/analyze.py Outdated Show resolved Hide resolved
config/base/config.yaml Outdated Show resolved Hide resolved
@alex404 alex404 requested a review from fabioseel October 24, 2024 09:35
fabioseel
fabioseel approved these changes Oct 24, 2024
View reviewed changes
Copy link
Contributor

@fabioseel fabioseel left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good now :)

@fabioseel fabioseel merged commit a80eac0 into master Oct 24, 2024
3 checks passed
@fabioseel fabioseel deleted the plot_control branch October 24, 2024 09:38
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@fabioseel fabioseel fabioseel approved these changes

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

Better analysis/plot scheduling
2 participants
@alex404 @fabioseel