mindee/doctr

Issues

• Cannot use builtin datasets for detection training

  #1830 opened 7 days ago by KenjiTakahashi
  1

• Builtin datasets get unpacked every time

  #1829 opened 14 days ago by KenjiTakahashi
  6

• unknown architecture: <class 'torch._dynamo.eval_frame.OptimizedModule'>

  #1819 opened a month ago by VMM-MMV
  2

• Modifying crop orientation prediction model classes does not result in appropriate rectification

  #1815 opened a month ago by milosacimovic
  5

• Getting incorrect page orientation value

  #1801 opened a month ago by vikasr111
  7

• [experimential] [detection] model training iteration with updated augmentation pipeline

  #1604 opened 8 months ago by felixdittrich92
  1

• python 3.13 compatibility issues (windows)

  #1814 opened a month ago by diegonc
  7

• Bug in multi-column page with dense text

  #1795 opened a month ago by vikasr111
  5

• Add support for torch.compile

  #1684 opened a month ago by Fabioomega
  13

• Release tracker - v0.11.0

  #1752 opened 3 months ago by felixdittrich92
  0

• OCR failure on certain PDF pages

  #1808 opened a month ago by vikasr111
  6

• ValueError: Layer 'conv2d_9' expected 1 variables, but received 0 variables during loading. Expected: ['conv2d_9/kernel:0']

  #1792 opened 2 months ago by MuhammadShifa
  10

• Multilingual support

  #1699 opened 5 months ago by decadance-dance
  28

• [Docker] Update Dockerfile to CUDA 12.X

  #1743 opened 2 months ago by felixdittrich92
  0

• PermissionError WinError 32, cannot access file and ParseError invalid token

  #1778 opened 2 months ago by zkfazal
  3

• How to get pixel location of detected bounding box?

  #1772 opened 2 months ago by DhiviJR
  0

• Inconsistency between `detection_predictor` and `ocr_predictor().det_predictor` resize parameters

  #1764 opened 2 months ago by cmoscardi
  2

• page.synthesize() returns error when assume_straight_pages=False

  #1692 opened 3 months ago by hienphan161
  2

• Release tracker - v0.10.0

  #1634 opened 3 months ago by felixdittrich92
  0

• Hindi Language support

  #1617 opened 3 months ago by chaudhary-mohit
  7

• [improvement] `.render()` isn't that robust - wrong ordered results

  #1586 opened 8 months ago by kripper
  12

• Release tracker - v0.9.1

  #1688 opened 3 months ago by felixdittrich92
  0

• [orientation] Make it possible to use your own trained orientation models

  #1706 opened 5 months ago by felixdittrich92
  0

• result.show() not displaying bounding boxes correctly

  #1691 opened 5 months ago by hienphan161
  5

• How to use doctr in Android and iOS?

  #1713 opened 3 months ago by yuvaraj119
  1

• UnboundLocalError: local variable 'l1_loss' referenced before assignment

  #1738 opened 3 months ago by johnlockejrr
  41

• Possible Issue with the current License

  #1730 opened 3 months ago by gvalvano
  3

• kwargs should pass through DocumentBuilder

  #1720 opened 4 months ago by TrevorDohm
  0

• ValueError: File format not supported: filepath=/root/.cache/doctr/models/parseq-24cf693e/weights. Keras 3 only supports V3 .keras and .weights.h5 files, or legacy V1/V2 .h5 files.

  #1719 opened 4 months ago by abhijeetsourav
  3

• Issue creating h5py wheel when running on M1 MacBook - trying to 'make run' in api directory

  #1683 opened 5 months ago by mamsterla
  9

• Add CTC Beam Search

  #1671 opened 6 months ago by ArsalanYounus007
  0

• UnboundLocalError: local variable 'l1_loss' referenced before assignment in Doctr's compute_loss function

  #1661 opened 6 months ago by ep0p
  2

• Weights not loaded for fast_tiny

  #1665 opened 6 months ago by ved-genmo
  1

• Optional dependencies get always installed

  #1658 opened 7 months ago by jonschill24
  2

• Removing noise in doctr

  #1655 opened 7 months ago by navaneeth-nagu
  0

• "The size of tensor a (78) must match the size of tensor b (77) at non-singleton dimension 3" with db_resnet* models

  #1650 opened 7 months ago by ForceBru
  1

• Add py.typed to the library

  #1638 opened 7 months ago by fzhem
  1

• Recognition training errors

  #1645 opened 7 months ago by ArsalanYounus007
  0

• -> Cannot close object, library is destroyed. This may cause a memory leak!

  #1623 opened 7 months ago by justinjosephmkj
  0

• `AttributeError: 'Sequential' object has no attribute 'get_output_at'` while using `fast_base` detection model.

  #1637 opened 7 months ago by Adversarian
  1

• OMG Look at my CPU memory usage increase insanely. Memory leak issue.

  #1594 opened 8 months ago by sanjay-nit
  4

• Words Detection Merged Bboxes

  #1622 opened 7 months ago by ArsalanYounus007
  8

• Partial detection when using fine-tuned model for OCR.

  #1614 opened 8 months ago by ashishupadhyaa
  0

• Missing dependencies

  #1589 opened 8 months ago by sylvainkalache
  2

• Can't run train_tensorflow.py on Google Colab GPU - only works on CPU

  #1591 opened 8 months ago by fosple
  4

• fine tune DBNET problem

  #1599 opened 8 months ago by SalehBM
  4

• Save the OCRed PDF

  #1595 opened 8 months ago by micos7
  2

• Tutorial export_as_pdfa is not working anymore on Colab

  #1596 opened 8 months ago by cthiounn
  0

• Reduce the build size for inference

  #1593 opened 8 months ago by decadance-dance
  4

• Missing depency libGL

  #1590 opened 8 months ago by sylvainkalache
  1