TypeError when using batch_layout_detection for multiple images

[ChernXiangyu]

Hi guys, I met this problem when trying to perform layout analysis in batches with the sample code in README. I guess this may because img_counts has become list of list when using multiple images as input.

I am running this code in Jupyter notebook in vscode, using M4Pro MacBook, macOS Squoia 15.2. Python 3.11.

Thanks in advance!

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
Cell In[17], line 16
     14 # layout_predictions is a list of dicts, one per image
     15 line_predictions = batch_text_detection(imgs[:2], det_model, det_processor)
---> 16 layout_predictions = batch_layout_detection(imgs[:2], model, processor, line_predictions)
     18 layout_predictions

File /opt/miniconda3/envs/fin-rag/lib/python3.11/site-packages/surya/layout.py:80, in batch_layout_detection(images, model, processor, batch_size, top_k)
     77 end_idx = 1
     78 while end_idx < len(img_counts):
     79     if any([
---> 80         sum(img_counts[start_idx:end_idx]) >= batch_size,
     81         sum(img_counts[start_idx:end_idx + 1]) > batch_size,
     82     ]):
     83         batches.append((start_idx, end_idx))
     84         start_idx = end_idx

TypeError: '>=' not supported between instances of 'int' and 'list'