.\.venv312\Scripts\python.exe : `torch_dtype` is deprecated! Use `dtype` instead!
At line:1 char:67
+ ... df" -Force; .\.venv312\Scripts\python.exe run_pipeline.py --pages 3 > ...
+                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    + CategoryInfo          : NotSpecified: (`torch_dtype` i...dtype` instead!:String) [], RemoteException
    + FullyQualifiedErrorId : NativeCommandError
 
Found file to process: test_full.pdf
Limiting processing to specific pages: [3]
Starting full pipeline extraction on input\test_full.pdf
Output directory: output\20260614_123256

Loaded detection model vikp/surya_layout3 on device cpu with dtype torch.float32

Detecting bboxes:   0%|          | 0/1 [00:00<?, ?it/s]
Detecting bboxes: 100%|ΓûêΓûêΓûêΓûêΓûêΓûêΓûêΓûêΓûêΓûê| 1/1 [00:11<00:00, 11.64s/it]
Detecting bboxes: 100%|ΓûêΓûêΓûêΓûêΓûêΓûêΓûêΓûêΓûêΓûê| 1/1 [00:11<00:00, 11.64s/it]

======================================================================
LAYOUT DETECTION RESULTS: 60 regions found
======================================================================
Region types: {'text': 41, 'paragraph_title': 12, 'image': 7}

  Region   1: type=image           bbox=[   88,  240, 1179,  745] size= 1091x  505px conf=1.00
  Region   2: type=image           bbox=[ 3075,  267, 4074,  867] size=  999x  600px conf=1.00
  Region   3: type=image           bbox=[ 1110,  284, 3029,  894] size= 1919x  610px conf=1.00
  Region   4: type=image           bbox=[ 1866, 1338, 2428, 1887] size=  562x  549px conf=1.00
  Region   5: type=image           bbox=[  582, 2916, 2079, 3815] size= 1497x  899px conf=1.00
  Region   6: type=image           bbox=[ 1570, 4147, 2074, 4536] size=  504x  389px conf=1.00
  Region   7: type=image           bbox=[  127, 4652, 3145, 5450] size= 3018x  798px conf=1.00
  Region   8: type=paragraph_title bbox=[  168, 1018, 1078, 1083] size=  910x   65px conf=1.00
  Region   9: type=paragraph_title bbox=[ 1246, 1019, 1792, 1081] size=  546x   62px conf=1.00
  Region  10: type=paragraph_title bbox=[  147, 1074,  607, 1133] size=  460x   59px conf=1.00
  Region  11: type=paragraph_title bbox=[  143, 1758, 1132, 1881] size=  989x  123px conf=1.00
  Region  12: type=paragraph_title bbox=[ 2118, 2592, 4026, 2736] size= 1908x  144px conf=1.00
  Region  13: type=paragraph_title bbox=[  184, 2593, 2008, 2791] size= 1824x  198px conf=1.00
  Region  14: type=paragraph_title bbox=[  471, 2787,  496, 2794] size=   25x    7px conf=1.00
  Region  15: type=paragraph_title bbox=[ 2787, 3353, 3387, 3421] size=  600x   68px conf=1.00
  Region  16: type=paragraph_title bbox=[  132, 6396,  625, 6471] size=  493x   75px conf=1.00
  Region  17: type=paragraph_title bbox=[ 1271, 6399, 1743, 6464] size=  472x   65px conf=1.00
  Region  18: type=paragraph_title bbox=[ 3596, 6398, 4052, 6467] size=  456x   69px conf=1.00
  Region  19: type=paragraph_title bbox=[ 2452, 6401, 2920, 6463] size=  468x   62px conf=1.00
  Region  20: type=text            bbox=[ 1829,  103, 2255,  204] size=  426x  101px conf=1.00
  Region  21: type=text            bbox=[ 3089,  114, 3784,  181] size=  695x   67px conf=1.00
  Region  22: type=text            bbox=[  198,  122,  759,  183] size=  561x   61px conf=1.00
  Region  23: type=text            bbox=[  237,  738, 1019,  802] size=  782x   64px conf=1.00
  Region  24: type=text            bbox=[ 2145,  809, 3121, 1028] size=  976x  219px conf=1.00
  Region  25: type=text            bbox=[  155,  833, 1107, 1035] size=  952x  202px conf=1.00
  Region  26: type=text            bbox=[ 3367,  860, 3781,  923] size=  414x   63px conf=1.00
  Region  27: type=text            bbox=[ 1252,  944, 1942, 1005] size=  690x   61px conf=1.00
  Region  28: type=text            bbox=[ 3080,  946, 4030, 1056] size=  950x  110px conf=1.00
  Region  29: type=text            bbox=[ 2149, 1011, 3125, 1922] size=  976x  911px conf=1.00
  Region  30: type=text            bbox=[ 1394, 1025, 1866, 1090] size=  472x   65px conf=1.00
  Region  31: type=text            bbox=[ 3125, 1062, 3281, 1128] size=  156x   66px conf=1.00
  Region  32: type=text            bbox=[ 1240, 1097, 2068, 1173] size=  828x   76px conf=1.00
  Region  33: type=text            bbox=[ 3076, 1115, 4072, 1921] size=  996x  806px conf=1.00
  Region  34: type=text            bbox=[   94, 1130, 1173, 1689] size= 1079x  559px conf=1.00
  Region  35: type=text            bbox=[ 1234, 1189, 2126, 1259] size=  892x   70px conf=1.00
  Region  36: type=text            bbox=[ 1203, 1288, 2272, 2481] size= 1069x 1193px conf=1.00
  Region  37: type=text            bbox=[  140, 1883,  889, 1947] size=  749x   64px conf=1.00
  Region  38: type=text            bbox=[   94, 1934, 1168, 2506] size= 1074x  572px conf=1.00
  Region  39: type=text            bbox=[ 2169, 1943, 4048, 2468] size= 1879x  525px conf=1.00
  Region  40: type=text            bbox=[ 2108, 2746, 2801, 3550] size=  693x  804px conf=1.00
  Region  41: type=text            bbox=[ 2756, 2745, 3449, 3319] size=  693x  574px conf=1.00
  Region  42: type=text            bbox=[ 3399, 2746, 4086, 3308] size=  687x  562px conf=1.00
  Region  43: type=text            bbox=[  209, 2821, 2001, 2888] size= 1792x   67px conf=1.00
  Region  44: type=text            bbox=[  135, 2916,  758, 4046] size=  623x 1130px conf=1.00
  Region  45: type=text            bbox=[  726, 3392, 1353, 4049] size=  627x  657px conf=1.00
  Region  46: type=text            bbox=[ 2744, 3412, 4092, 4032] size= 1348x  620px conf=1.00
  Region  47: type=text            bbox=[ 2141, 3547, 2460, 3606] size=  319x   59px conf=1.00
  Region  48: type=text            bbox=[ 2108, 3599, 2797, 4061] size=  689x  462px conf=1.00
  Region  49: type=text            bbox=[ 1320, 3829, 2050, 4045] size=  730x  216px conf=1.00
  Region  50: type=text            bbox=[  326, 4034, 1294, 4574] size=  968x  540px conf=1.00
  Region  51: type=text            bbox=[ 1142, 4067, 1962, 4503] size=  820x  436px conf=1.00
  Region  52: type=text            bbox=[ 3076, 4082, 4093, 4606] size= 1017x  524px conf=1.00
  Region  53: type=text            bbox=[ 2091, 4107, 3119, 4599] size= 1028x  492px conf=1.00
  Region  54: type=text            bbox=[ 3240, 4678, 4019, 4927] size=  779x  249px conf=1.00
  Region  55: type=text            bbox=[ 3196, 4917, 4077, 6365] size=  881x 1448px conf=1.00
  Region  56: type=text            bbox=[ 2350, 4999, 3145, 6366] size=  795x 1367px conf=1.00
  Region  57: type=text            bbox=[  127, 5227,  926, 6372] size=  799x 1145px conf=1.00
  Region  58: type=text            bbox=[ 1149, 5441, 2113, 5506] size=  964x   65px conf=1.00
  Region  59: type=text            bbox=[  866, 5522, 1661, 6363] size=  795x  841px conf=1.00
  Region  60: type=text            bbox=[ 1611, 5522, 2406, 6369] size=  795x  847px conf=1.00
======================================================================

DEBUG IMAGE saved: output\20260614_123256\pages\page_003_regions_debug.png
  Open this image to see all 60 bounding boxes drawn on the page
  Colors: text=blue, title/doc_title=red, image/figure=green, header/footer=gray, paragraph_title=orange

  Page 3: 60 layout regions detected
[running geometric column grouping]
  [split] Article 5: split at gap=124px after region 39 (upper 4 regions, lower 6 regions)
  [split] Article 8: split at gap=72px after region 53 (upper 4 regions, lower 2 regions)

======================================================================
GROUPING RESULTS ΓÇö PAGE 3
======================================================================
  Input: 60 regions -> Output: 15 articles

  ARTICLE 1: ""
    Title region: ID 8 (type=paragraph_title, bbox=[168, 1018, 1078, 1083])
    Member regions (6): [8, 1, 22, 23, 25, 34]
      Region   8: type=paragraph_title bbox=[  168, 1018, 1078, 1083] size=  910x   65px
      Region   1: type=image           bbox=[   88,  240, 1179,  745] size= 1091x  505px
      Region  22: type=text            bbox=[  198,  122,  759,  183] size=  561x   61px
      Region  23: type=text            bbox=[  237,  738, 1019,  802] size=  782x   64px
      Region  25: type=text            bbox=[  155,  833, 1107, 1035] size=  952x  202px
      Region  34: type=text            bbox=[   94, 1130, 1173, 1689] size= 1079x  559px
    Merged bbox: [88,122,1179,1689] size=1091x1567px
    Article bbox: [88, 122, 1179, 1689]

  ARTICLE 2: ""
    Title region: ID 9 (type=paragraph_title, bbox=[1246, 1019, 1792, 1081])
    Member regions (9): [9, 3, 20, 24, 27, 30, 32, 35, 36]
      Region   9: type=paragraph_title bbox=[ 1246, 1019, 1792, 1081] size=  546x   62px
      Region   3: type=image           bbox=[ 1110,  284, 3029,  894] size= 1919x  610px
      Region  20: type=text            bbox=[ 1829,  103, 2255,  204] size=  426x  101px
      Region  24: type=text            bbox=[ 2145,  809, 3121, 1028] size=  976x  219px
      Region  27: type=text            bbox=[ 1252,  944, 1942, 1005] size=  690x   61px
      Region  30: type=text            bbox=[ 1394, 1025, 1866, 1090] size=  472x   65px
Traceback (most recent call last):
      Region  32: type=text            bbox=[ 1240, 1097, 2068, 1173] size=  828x   76px
      Region  35: type=text            bbox=[ 1234, 1189, 2126, 1259] size=  892x   70px
      Region  36: type=text            bbox=[ 1203, 1288, 2272, 2481] size= 1069x 1193px
    Merged bbox: [1110,103,3121,2481] size=2011x2378px
    Article bbox: [1110, 103, 3121, 2481]

  ARTICLE 3: ""
    Title region: ID 10 (type=paragraph_title, bbox=[147, 1074, 607, 1133])
    Member regions (1): [10]
      Region  10: type=paragraph_title bbox=[  147, 1074,  607, 1133] size=  460x   59px
    Merged bbox: [147,1074,607,1133] size=460x59px
    Article bbox: [147, 1074, 607, 1133]

  ARTICLE 4: ""
    Title region: ID 11 (type=paragraph_title, bbox=[143, 1758, 1132, 1881])
    Member regions (3): [11, 37, 38]
      Region  11: type=paragraph_title bbox=[  143, 1758, 1132, 1881] size=  989x  123px
      Region  37: type=text            bbox=[  140, 1883,  889, 1947] size=  749x   64px
      Region  38: type=text            bbox=[   94, 1934, 1168, 2506] size= 1074x  572px
    Merged bbox: [94,1758,1168,2506] size=1074x748px
    Article bbox: [94, 1758, 1168, 2506]

  ARTICLE 5: ""
    Title region: ID 12 (type=paragraph_title, bbox=[2118, 2592, 4026, 2736])
    Member regions (4): [29, 33, 4, 39]
      Region  29: type=text            bbox=[ 2149, 1011, 3125, 1922] size=  976x  911px
      Region  33: type=text            bbox=[ 3076, 1115, 4072, 1921] size=  996x  806px
      Region   4: type=image           bbox=[ 1866, 1338, 2428, 1887] size=  562x  549px
      Region  39: type=text            bbox=[ 2169, 1943, 4048, 2468] size= 1879x  525px
    Merged bbox: [1866,1011,4072,2468] size=2206x1457px
    Article bbox: [1866, 1011, 4072, 2468]

  ARTICLE 6: ""
    Member regions (6): [12, 41, 40, 42, 47, 48]
      Region  12: type=paragraph_title bbox=[ 2118, 2592, 4026, 2736] size= 1908x  144px
      Region  41: type=text            bbox=[ 2756, 2745, 3449, 3319] size=  693x  574px
      Region  40: type=text            bbox=[ 2108, 2746, 2801, 3550] size=  693x  804px
      Region  42: type=text            bbox=[ 3399, 2746, 4086, 3308] size=  687x  562px
      Region  47: type=text            bbox=[ 2141, 3547, 2460, 3606] size=  319x   59px
      Region  48: type=text            bbox=[ 2108, 3599, 2797, 4061] size=  689x  462px
    Merged bbox: [2108,2592,4086,4061] size=1978x1469px
    Article bbox: [2108, 2592, 4086, 4061]

  ARTICLE 7: ""
    Title region: ID 13 (type=paragraph_title, bbox=[184, 2593, 2008, 2791])
    Member regions (9): [13, 5, 6, 7, 43, 45, 49, 50, 51]
      Region  13: type=paragraph_title bbox=[  184, 2593, 2008, 2791] size= 1824x  198px
      Region   5: type=image           bbox=[  582, 2916, 2079, 3815] size= 1497x  899px
      Region   6: type=image           bbox=[ 1570, 4147, 2074, 4536] size=  504x  389px
      Region   7: type=image           bbox=[  127, 4652, 3145, 5450] size= 3018x  798px
      Region  43: type=text            bbox=[  209, 2821, 2001, 2888] size= 1792x   67px
      Region  45: type=text            bbox=[  726, 3392, 1353, 4049] size=  627x  657px
      Region  49: type=text            bbox=[ 1320, 3829, 2050, 4045] size=  730x  216px
      Region  50: type=text            bbox=[  326, 4034, 1294, 4574] size=  968x  540px
      Region  51: type=text            bbox=[ 1142, 4067, 1962, 4503] size=  820x  436px
    Merged bbox: [127,2593,3145,5450] size=3018x2857px
    Article bbox: [127, 2593, 3145, 5450]

  ARTICLE 8: ""
    Title region: ID 14 (type=paragraph_title, bbox=[471, 2787, 496, 2794])
    Member regions (2): [14, 44]
      Region  14: type=paragraph_title bbox=[  471, 2787,  496, 2794] size=   25x    7px
      Region  44: type=text            bbox=[  135, 2916,  758, 4046] size=  623x 1130px
    Merged bbox: [135,2787,758,4046] size=623x1259px
    Article bbox: [135, 2787, 758, 4046]

  ARTICLE 9: ""
    Title region: ID 15 (type=paragraph_title, bbox=[2787, 3353, 3387, 3421])
    Member regions (4): [15, 46, 52, 53]
      Region  15: type=paragraph_title bbox=[ 2787, 3353, 3387, 3421] size=  600x   68px
      Region  46: type=text            bbox=[ 2744, 3412, 4092, 4032] size= 1348x  620px
      Region  52: type=text            bbox=[ 3076, 4082, 4093, 4606] size= 1017x  524px
      Region  53: type=text            bbox=[ 2091, 4107, 3119, 4599] size= 1028x  492px
    Merged bbox: [2091,3353,4093,4606] size=2002x1253px
    Article bbox: [2091, 3353, 4093, 4606]

  ARTICLE 10: ""
    Member regions (2): [54, 55]
      Region  54: type=text            bbox=[ 3240, 4678, 4019, 4927] size=  779x  249px
      Region  55: type=text            bbox=[ 3196, 4917, 4077, 6365] size=  881x 1448px
    Merged bbox: [3196,4678,4077,6365] size=881x1687px
    Article bbox: [3196, 4678, 4077, 6365]

  ARTICLE 11: ""
    Title region: ID 16 (type=paragraph_title, bbox=[132, 6396, 625, 6471])
    Member regions (2): [16, 57]
      Region  16: type=paragraph_title bbox=[  132, 6396,  625, 6471] size=  493x   75px
      Region  57: type=text            bbox=[  127, 5227,  926, 6372] size=  799x 1145px
    Merged bbox: [127,5227,926,6471] size=799x1244px
    Article bbox: [127, 5227, 926, 6471]

  ARTICLE 12: ""
    Title region: ID 17 (type=paragraph_title, bbox=[1271, 6399, 1743, 6464])
    Member regions (4): [17, 58, 59, 60]
      Region  17: type=paragraph_title bbox=[ 1271, 6399, 1743, 6464] size=  472x   65px
      Region  58: type=text            bbox=[ 1149, 5441, 2113, 5506] size=  964x   65px
      Region  59: type=text            bbox=[  866, 5522, 1661, 6363] size=  795x  841px
      Region  60: type=text            bbox=[ 1611, 5522, 2406, 6369] size=  795x  847px
    Merged bbox: [866,5441,2406,6464] size=1540x1023px
    Article bbox: [866, 5441, 2406, 6464]

  ARTICLE 13: ""
    Title region: ID 18 (type=paragraph_title, bbox=[3596, 6398, 4052, 6467])
    Member regions (1): [18]
      Region  18: type=paragraph_title bbox=[ 3596, 6398, 4052, 6467] size=  456x   69px
    Merged bbox: [3596,6398,4052,6467] size=456x69px
    Article bbox: [3596, 6398, 4052, 6467]

  ARTICLE 14: ""
    Title region: ID 19 (type=paragraph_title, bbox=[2452, 6401, 2920, 6463])
    Member regions (2): [19, 56]
      Region  19: type=paragraph_title bbox=[ 2452, 6401, 2920, 6463] size=  468x   62px
      Region  56: type=text            bbox=[ 2350, 4999, 3145, 6366] size=  795x 1367px
    Merged bbox: [2350,4999,3145,6463] size=795x1464px
    Article bbox: [2350, 4999, 3145, 6463]

  ARTICLE 15: ""
    Member regions (5): [21, 2, 26, 28, 31]
      Region  21: type=text            bbox=[ 3089,  114, 3784,  181] size=  695x   67px
      Region   2: type=image           bbox=[ 3075,  267, 4074,  867] size=  999x  600px
      Region  26: type=text            bbox=[ 3367,  860, 3781,  923] size=  414x   63px
      Region  28: type=text            bbox=[ 3080,  946, 4030, 1056] size=  950x  110px
      Region  31: type=text            bbox=[ 3125, 1062, 3281, 1128] size=  156x   66px
    Merged bbox: [3075,114,4074,1128] size=999x1014px
    Article bbox: [3075, 114, 4074, 1128]

  STATS:
    Articles: 15
    Regions per article: min=1, max=9, avg=4.0
    ⚠️ Large articles (>6 regions): [2, 7] — may have merged multiple stories
======================================================================

ARTICLES DEBUG IMAGE saved: output\20260614_123256\pages\page_003_articles_debug.png
  Each article has a unique color. Regions with same color = same article.

  Page 3: filtered out 1 tiny/garbage articles (< 60px or < 10000px┬▓), 14 remain for cropping.
  File "C:\Users\sunde\proxmox\sundeep-news-scan\run_pipeline.py", line 107, in <module>
    main()
  File "C:\Users\sunde\proxmox\sundeep-news-scan\run_pipeline.py", line 77, in main
    result = extractor.process_pdf(
             ^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\sunde\proxmox\sundeep-news-scan\extractor.py", line 887, in process_pdf
    headline_records = crop_headlines(page_img_path, articles, regions, articles_root, n)
                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\sunde\proxmox\sundeep-news-scan\extractor.py", line 253, in crop_headlines
    crop = img.crop((l, t, r, b))
           ^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\sunde\proxmox\sundeep-news-scan\.venv312\Lib\site-packages\PIL\Image.py", line 1299, in crop
    raise ValueError(msg)
ValueError: Coordinate 'lower' is less than 'upper'