myhloli
|
fcf24242d1
chore: remove useless files
|
1 year ago |
myhloli
|
564c4ce1e3
refactor(magic_pdf): improve line sorting and block indexing
|
1 year ago |
myhloli
|
4c9bf8abd5
refactor(memory management): remove unused clean_memory function
|
1 year ago |
myhloli
|
42a7d792c3
refactor(magic_pdf): import model helpers directly for clarity
|
1 year ago |
myhloli
|
5522d0a36c
refactor(pdf_parse_union_core_v2): update import paths to use new package structure
|
1 year ago |
myhloli
|
2145a8b6d2
fix(pdf_parse): handle blocks without lines and enable bf16 on compatible devices
|
1 year ago |
myhloli
|
177ab08e9f
refactor(pdf_parse): remove redundant sorting and optimize block indexing
|
1 year ago |
myhloli
|
83c073875a
refactor(draw_bbox): remove commented-out code and streamline bbox drawingRemoved legacy commented-out code related to layout_bbox_list from draw_bbox.py, which
|
1 year ago |
myhloli
|
656154558a
feat(requirements): add torch and transformers libraries
|
1 year ago |
myhloli
|
b9dfdea3cb
refactor(pdf_parse_union_core_v2): implement model initialization within classRefactored model initialization to be handled by a singleton class to ensure that model
|
1 year ago |
myhloli
|
b2790f6f45
refactor(drawing): simplify draw bbox functions and adjust debug config
|
1 year ago |
myhloli
|
16b51c79c8
Merge remote-tracking branch 'origin/add-layoutreader' into add-layoutreader
|
1 year ago |
myhloli
|
43a57d5627
feat(draw_bbox): add option to toggle bounding box drawing
|
1 year ago |
myhloli
|
c56de49377
refactor(draw_bbox): remove conditional layout bbox drawing
|
1 year ago |
myhloli
|
34f8965007
refactor(draw_bbox): add line sorting visualization
|
1 year ago |
myhloli
|
1efebe421c
refactor(pdf_parse_union): integrate LayoutLMv3 for block orderingReplace the heuristic-based block ordering algorithm with LayoutLMv3 model predictions toimprove the accuracy of block ordering on PDF pages. Additionally, refactor the span
|
1 year ago |
myhloli
|
36220d69d4
refactor(draw_bbox): clear cuda cache and update bbox sorting
|
1 year ago |
myhloli
|
00cda7a6f4
refactor(draw_bbox): clear cuda cache and update bbox sorting
|
1 year ago |
myhloli
|
270ffb020d
feat(draw_bbox): add layout sorting visualization
|
1 year ago |
myhloli
|
3cbcf2ded0
feat(draw_bbox): add layout sorting visualization
|
1 year ago |
Xiaomeng Zhao
|
24c143fe8d
Merge pull request #639 from myhloli/dev
|
1 year ago |
myhloli
|
f2a3a49541
fix(pdf_extract_kit):change unimernet base -> small
|
1 year ago |
Xiaomeng Zhao
|
3a616cf29d
Merge pull request #636 from myhloli/dev
|
1 year ago |
myhloli
|
4811a3d1df
fix(pdf-extract): ensure model is set to evaluation mode before processing
|
1 year ago |
Xiaomeng Zhao
|
19a74db869
Merge pull request #635 from myhloli/dev
|
1 year ago |
myhloli
|
c36fa049b0
refactor(pdf_extract): use Image.crop directly with layout detection
|
1 year ago |
Xiaomeng Zhao
|
fb383ba668
Merge pull request #631 from myhloli/dev
|
1 year ago |
myhloli
|
23b621e05a
feat(UNIPipe): change default drop_mode to NONE_WITH_REASON
|
1 year ago |
Xiaomeng Zhao
|
c4a52ee6c6
Merge pull request #630 from myhloli/content-list-not-drop
|
1 year ago |
Xiaomeng Zhao
|
98313d4a25
Merge branch 'dev' into content-list-not-drop
|
1 year ago |