Skip to content

Revamp overall report UI and improve report navigation #1102

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

Open
wants to merge 10 commits into
base: main
Choose a base branch
from

Conversation

myanvoos
Copy link
Collaborator

@myanvoos myanvoos commented Jun 8, 2025

This PR overhauls the overall report UI (layout and template/sample pages) and adds a few improvements including:

  • A dark mode theme for all three pages
  • Navigation breadcrumbs
  • Toggle-able table of content for the index page (located top right of the screen)
  • Dedicated header showing aggregated results for the index and benchmark pages
  • A function to create a unified JSON file out of experimental results that preserves the hierarchical relationships between projects, benchmarks, and samples

@myanvoos myanvoos self-assigned this Jun 8, 2025
Copy link
Collaborator

@DavidKorczynski DavidKorczynski left a comment

Choose a reason for hiding this comment

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

Could you post screenshots?

IMO on thing that we should watch out for is losing overview ability for the benefit of more modern-looking UI. Following #1062 I think it's more difficult to traverse the benchmarks on a higher level, probably because we have significantly reduced the amount of benchmarks shown on the screen at once. I find when experiments have hundreds of benchmarks it's convenient to have a lot of data shown at once.

@myanvoos
Copy link
Collaborator Author

Could you post screenshots?

IMO on thing that we should watch out for is losing overview ability for the benefit of more modern-looking UI. Following #1062 I think it's more difficult to traverse the benchmarks on a higher level, probably because we have significantly reduced the amount of benchmarks shown on the screen at once. I find when experiments have hundreds of benchmarks it's convenient to have a lot of data shown at once.

Thank you for your feedback @DavidKorczynski!

Do you think making all benchmarks shown as open by default would help in this regard? Currently their default behaviour is collapsed unless you clicked the 'Expand All' button or their corresponding project.

Alternatively, would a dedicated table for benchmarks without grouping to projects, similar to the old report, be more helpful? It can exist alongside the current nested 'Project Summary' tables, like this:

image

Here are the screenshots:

Index page without table of content opened

image

Index page with table of content opened

image

Index page in dark mode, with table of content opened

image

Benchmark page with prompt breakdown accordion opened

image

Benchmark page in dark mode, with prompt breakdown accordion collapsed

image

Trial/Sample page upper half

image

Trial/Sample page lower half, with final code, logs accordions, and the embedded coverage report

image

@DavidKorczynski
Copy link
Collaborator

The UI looks great, thanks for sharing the screenshots.

Do you think making all benchmarks shown as open by default would help in this regard? Currently their default behaviour is collapsed unless you clicked the 'Expand All' button or their corresponding projec

The default is okay as is now.

Alternatively, would a dedicated table for benchmarks without grouping to projects, similar to the old report, be more helpful? It can exist alongside the current nested 'Project Summary' tables, like this:

Yeah I think this is what we need. A table (can be on a separate page or w/e, just that it's accessible) with minimal whitespace and ability to see a lot of benchmarks at once, similar to previously.

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

Successfully merging this pull request may close these issues.

2 participants