[PeterStaar-IBM]

@wittjeff : please refrain from doing PR's for now. I am trying to measure (and hopefully improve) the overall performance of the docling-parse.

Evidence:

on

./build/run_scaling.exe /Users/taa/Documents/projects/_data/bo767/pdf \
  --mode render --threads 1,4,8,12 \
  --max-concurrent-results 32 --scale 1 \
  --create-word-cells=true --create-line-cells=true

we have so far,

backend               threads     wall_time (s)     vs threaded(1)     efficiency     pages/sec     ms/page    errors
------------------------------------------------------------------------------------------------------------------
docling threaded           1          1135.209             1.00x          100%          48.2       20.74         0
docling threaded           4           299.088             3.80x           95%         183.0        5.46         0
docling threaded           8           154.937             7.33x           92%         353.2        2.83         0
docling threaded          12           108.805            10.43x           87%         503.0        1.99         0

Ground truth changes:

Checklist:

• One topic per pull request, rebased on main, no merge commits
• Superseded pull requests closed and linked
• Regression dataset documents and pages named, or a dataset pull request that adds a new document
• Before/after evidence for each cited page
• Rendering only: pypdfium2 comparison images, corpus-wide metrics table, and an enforced image regression test
• Blend2D only: link to the upstream report
• Changes to existing ground truth listed with the reason for each page

[github-actions]

✅ DCO Check Passed

Thanks @PeterStaar-IBM, all your commits are properly signed off. 🎉

[mergify]

Merge Protections

🟢 Merge protection satisfied — ready to merge.

Show 1 satisfied protection

🟢 Enforce conventional commit

Make sure that we follow https://www.conventionalcommits.org/en/v1.0.0/

• title ~= ^(fix|feat|docs|style|refactor|perf|test|build|ci|chore|revert)(?:\(.+\))?(!)?: