Commit Graph

353 Commits

Author SHA1 Message Date
Vik Paruchuri
72d00647a5 Refactor benchmarks
Some checks failed
Integration test / build (push) Has been cancelled
Unit tests / build (push) Has been cancelled
2025-01-10 21:31:24 -05:00
Vik Paruchuri
03e722f613 Update model checkpoint 2025-01-10 14:34:40 -05:00
Vik Paruchuri
05b6ed7e6c Update benchmarks 2025-01-10 12:40:35 -05:00
Vik Paruchuri
7174903d54 Expand table boxes slightly 2025-01-10 09:10:23 -05:00
Vik Paruchuri
d2a03524e7 Add model support for headers
Some checks failed
Integration test / build (push) Has been cancelled
Unit tests / build (push) Has been cancelled
2025-01-08 21:32:56 -05:00
Vik Paruchuri
a1f500cc14 Fix minor issues 2025-01-08 15:28:22 -05:00
Vik Paruchuri
29bd086109 Remove some imports 2025-01-08 12:12:17 -05:00
Vik Paruchuri
bf2f693daa Additional cleanup 2025-01-08 12:10:32 -05:00
Vik Paruchuri
bdc244e573 Refactor CLI scripts 2025-01-07 19:56:39 -05:00
Vik Paruchuri
3cf4d297e1 Update README 2025-01-07 16:51:44 -05:00
Vik Paruchuri
840c7ab3cb Fix predictions 2025-01-07 16:47:40 -05:00
Vik Paruchuri
4c5a1807da Refactor batch sizes 2025-01-07 10:25:27 -05:00
Vik Paruchuri
d17281ea1a Refactor schema 2025-01-07 09:52:42 -05:00
Vik Paruchuri
dd667d27c6 Update benchmarks 2025-01-06 22:09:15 -05:00
Vik Paruchuri
8c3eb5f6ce Fix benchmarks 2025-01-06 22:04:07 -05:00
Vik Paruchuri
85301313fe Refactor table rec 2025-01-06 21:58:35 -05:00
Vik Paruchuri
51c7c4acc6 Refactor OCR error model 2025-01-06 21:39:45 -05:00
Vik Paruchuri
d7f1567346 Refactor layout 2025-01-06 17:38:51 -05:00
Vik Paruchuri
1de6d9193d Refactor recognition model 2025-01-06 17:19:17 -05:00
Vik Paruchuri
187fc8f2bc Refactor detection model 2025-01-06 15:08:19 -05:00
Vik Paruchuri
e048733043 Merge dev 2025-01-06 14:23:20 -05:00
Vik Paruchuri
10d353ea23 Fix poetry lock
Some checks failed
Integration test / build (push) Has been cancelled
2024-12-31 21:52:42 -05:00
Vik Paruchuri
66261ffce8 Pin pypdfium2 2024-12-30 15:53:49 -05:00
Vik Paruchuri
ffd3f5f2df New layout model
Some checks failed
Integration test / build (push) Has been cancelled
2024-12-30 12:05:36 -05:00
Vik Paruchuri
76754bca53 Update layout model
Some checks failed
Integration test / build (push) Has been cancelled
2024-12-20 14:56:07 -05:00
Vik Paruchuri
6bbdc57b0d Refactor encoder
Some checks failed
Integration test / build (push) Has been cancelled
2024-12-19 11:56:40 -05:00
Vik Paruchuri
4e60cc5e87 Add test for topk 2024-12-19 11:43:06 -05:00
Vik Paruchuri
cd795a71c0 Add in tests 2024-12-19 11:15:03 -05:00
Vik Paruchuri
2281aec8b9 Bump version and lockfile 2024-12-19 11:01:58 -05:00
Vik Paruchuri
20b7b62d5d Add bad OCR detection to app 2024-12-19 10:49:25 -05:00
Vik Paruchuri
f8188f41f4 Modify prediction logic
Some checks failed
Integration test / build (push) Has been cancelled
2024-12-18 14:11:38 -05:00
Vik Paruchuri
2525ee0ea1
Merge pull request #263 from VikParuchuri/dev-mose/layout_top_k
Add `top_k` to Surya Layout and Fix Confidence Value Issue
2024-12-18 09:28:24 -08:00
Vik Paruchuri
51dde0ecce
Merge pull request #261 from tarun-menta/ocr-error-model
Add OCR Error Detection Model
2024-12-18 09:27:15 -08:00
Tarun Menta
f63a4bf835
Update settings.py 2024-12-15 20:21:00 +05:30
Tarun Ram
872c31faf3 Minor bugfix 2024-12-14 21:46:23 +05:30
Tarun Ram
9f7ea2938a Add OCR Error Detection Model 2024-12-14 21:33:04 +05:30
Vik Paruchuri
4701f96de9 Fix bug in decoder
Some checks failed
Integration test / build (push) Has been cancelled
2024-12-13 16:58:53 -05:00
Moses Paul R
f000187f1d
fix missing confidence in surya layout predictions 2024-12-13 17:37:47 +00:00
Moses Paul R
d3cf58ae98
ditch blanks [skip ci] 2024-12-13 16:31:36 +00:00
Moses Paul R
ed05c016f5
proper probs 2024-12-13 16:02:07 +00:00
Moses Paul R
cf7ee06ef6
add top_k to layout results 2024-12-13 15:48:13 +00:00
Vik Paruchuri
af89a06eda Inference loop 2024-12-12 21:39:37 -05:00
Vik Paruchuri
d41dd7c839 Patch issues with table rec 2024-12-12 16:40:30 -05:00
Vik Paruchuri
237592926e Start to redo table recognition 2024-12-12 15:10:15 -05:00
Vik Paruchuri
573b762b17 Start to pull in new table model 2024-12-12 11:56:27 -05:00
Moses Paul R
b46d5ce2f6
Merge pull request #259 from VikParuchuri/dev
Some checks failed
Integration test / build (push) Has been cancelled
Bugfixes and `pdftext` improvements
2024-12-12 20:37:43 +04:00
Moses Paul R
a3fde2f7cf
fix readme bugs [skip ci] 2024-12-12 16:36:43 +00:00
Moses Paul R
0230848679
increment surya version
Some checks failed
Integration test / build (push) Has been cancelled
2024-12-12 16:21:32 +00:00
Moses Paul R
b42800b45c
increment surya and pdftext versions 2024-12-12 16:17:57 +00:00
Vik Paruchuri
128e5fbc7f Fix residual flow 2024-12-12 10:42:44 -05:00