GitLab

- Update batch ratio calculation for GPU memory range
- Increase upper limit for batch ratio 16 from 24 to 32 GB

- Add timing measurement for formula, text, and title optimization using LLM
- Log the execution time for each LLM aided process

- Add a check to return 0 when either bbox1_area or bbox2_area is zero
- This prevents division by zero errors when calculating IoU

- Restore commented code for filtering out characters with invalid bounding boxes
- This change may affect the filtering of unnecessary characters in PDF parsing

- Update conditions for batch ratio assignment:
  -8 <= gpu_memory < 10: batch_ratio = 2 - 10 <= gpu_memory <= 12: batch_ratio =4
- This fix ensures proper batch ratio selection for GPU memory sizes

- Improve batch ratio calculation based on GPU memory
- Enhance performance for devices with 8GB or more VRAM

- Reduce batch_ratio by 1 for better performance and stability
- This change ensures more consistent memory usage when processing documents

refactor(magic_pdf): adjust VRAM allocation and MFR batch size- Update VRAM allocation logic to use 'VIRTUAL_VRAM_SIZE' environment variable
- Reduce MFR (Math Formula Recognition) batch size from 64 to 32

- Update GPU memory check and batch ratio calculation logic
- Add support for virtual VRAM size environment variable
- Improve logging for GPU memory and batch ratio

- Reduce YOLO_LAYOUT_BASE_BATCH_SIZE from 4 to 1
- Simplify batch ratio calculation for formula detection
- Remove unused conditional logic in batch ratio determination

- Update model path from 'unimernet_small' to 'unimernet_small_2501' in multiple scripts and configuration files
- This change affects download_models.py, download_models_hf.py, and model_configs.yaml

- Add key length validation for ONNX model initialization
- Move import statements to the top of the file
- Wrap model initialization in a try-except block for better error handling
- Refactor code to improve readability and maintainability

- Add remove_tilted_line function to filter out lines with angles between 2 and 88 degrees
- Integrate the new function into the text extraction process
- Improve the accuracy of text block processing by removing non-horizontal/vertical lines

- Added instructions for checking the reasonability of heading levels
- Included guidelines for making fine adjustments based on context and logic
- Emphasized the importance of aligning the final result with the document's actual structure

- Commented out the original batch ratio calculation
- Set a fixed batch ratio of 2 for GPUs with less than 8 GB memory
- Increased batch ratio to 4 for GPUs with 8 GB or more memory

- Import get_device function from magic_pdf.libs.config_reader- Update RapidTableModel initialization to include device parameter for Unitable model

- Modify the batch analyze process to handle the rapid table model's output
- Add logic_points variable to capture additional output from rapid table prediction

- Update RapidTable dependency to version 1.0.3
- Add support for sub-models in RapidTable
- Update magic-pdf configuration to include table sub-model
- Modify table model initialization to support sub-models
- Update table prediction logic to handle new output format

- Adjust end_page_id calculation to prevent IndexError when accessing pages
- Enhance error handling in LLM post-processing by specifically catching JSONDecodeError

- Rename and update merge_title_blocks function
- Implement merge_two_bbox helper function
- Refactor merging logic to preserve original block structure- Update function calls and integrate with existing pipeline

- Add support for NPU (Neural Processing Unit) when available
- Implement batch analysis for GPU and NPU devices
- Optimize memory usage and improve performance
- Update logging and error handling

- Add `remove_invalid_surrogates` function to filter out invalid UTF-16 surrogate pairs
- Integrate the new function into the `detect_lang` workflow
- Include a test case with UTF-16 surrogates to verify the fix

- Clarify the expected format for the optimized title list JSON output- Emphasize the need to return only the title levels in the specified format

- Modified the IOU threshold in ocr_span_list_modify.py from 0.9 to 0.35
- This change aims to improve the detection of overlapping characters in OCR processed PDFs

- Add average line height calculation for title blocks
- Include page number in title dictionary
- Improve title optimization prompt for better hierarchy- Implement retry mechanism for JSON decoding errors
- Add error logging for title count mismatch

- Merge title blocks that are close to each other horizontally
- Adjust line insertion logic for title blocks- Increase image size and decrease confidence threshold for layout detection
- Update DocLayoutYOLO model weights
- Refactor drawing of bounding boxes for different block types

- Add enable flag check for formula, text, and title optimizations

- Add MPS support for Apple Silicon devices
- Implement empty_cache() for MPS devices
- Set PYTORCH_ENABLE_MPS_FALLBACK environment variable
- Adjust MFR model device allocation for MPS

- Improve language detection by removing newline characters from the input text
- Add error handling and fallback mechanism to deal with text containing control characters

- Remove conditional logic for OCR engine selection
- Always use RapidOCR as the OCR engine
- Simplify the __init__ method by removing unused code

- Remove YOLO v11 language detection model from model_configs.yaml
- Update language detection utils to use a fixed model path instead of dynamic configuration
- Remove unused model weight parameter for YOLO v11 language detection

- Implement block sorting within image and table blocks
- Ensure correct order of captions and footnotes within blocks
- Improve overall document structure and parsing accuracy

- Remove LangDetectMode and related conditional logic
- Use a single model weight for language detection
- Add logging for language detection results
- Update model initialization and prediction methods

- Add language detection model initialization and integration
- Update model list to include language detection
- Refactor language detection utils for better model management