Коммит түүх

Эзэн SHA1 Мессеж Огноо
  zhch158_admin b6a4b45c91 feat: implement financial number normalization across output formats, enhancing data consistency and accuracy in JSON, Markdown, and HTML exports 1 долоо хоног өмнө
  zhch158_admin db78f28b68 feat: integrate DoclingLayoutDetector into the adapter module, expanding layout detection capabilities and updating the export list for improved accessibility 1 долоо хоног өмнө
  zhch158_admin e14a1607c7 refactor: update category mapping and color definitions in PaddleLayoutDetector for improved clarity and consistency with EnhancedDocPipeline 1 долоо хоног өмнө
  zhch158_admin f98e8cc99f feat: implement DoclingLayoutDetector adapter for layout detection using HuggingFace models, supporting batch processing and visualization, while ensuring compatibility with existing MinerU categories 1 долоо хоног өмнө
  zhch158_admin 22f3d967cf feat: add support for Docling layout detector in ModelFactory, enhancing layout detection options and improving error handling during initialization 1 долоо хоног өмнө
  zhch158_admin edd2eea7ce feat: add enhanced bank statement configuration files for MinerU and PaddleOCR-VL, supporting improved processing workflows and output options, while removing the outdated financial report configuration 1 долоо хоног өмнө
  zhch158_admin c813baaef7 refactor: remove main_enhanced.py to streamline document processing scripts, consolidating functionality into main_v2.py for improved maintainability and clarity 1 долоо хоног өмнө
  zhch158_admin 7d96321bc1 feat: introduce new utility modules for document processing, including HTML and JSON formatters, along with a Markdown generator and visualization tools, enhancing output capabilities and modularity 1 долоо хоног өмнө
  zhch158_admin 0f448107b5 feat: add DocLayoutYOLOModel for document layout prediction, including batch processing and visualization capabilities, along with a new test image for validation 1 долоо хоног өмнө
  zhch158_admin 57c61906b5 feat: add core modules for document processing including coordinate utilities, element processors, layout utilities, and PDF handling, enhancing overall functionality and structure 1 долоо хоног өмнө
  zhch158_admin a5b6546c7d fix: update installation instructions for the merger module to ensure proper uninstallation and reinstallation 1 долоо хоног өмнө
  zhch158_admin d86fb0815b refactor: remove MinerUOutputFormatter and related classes, streamline output formatting in OutputFormatterV2, and enhance JSON and Markdown generation capabilities 1 долоо хоног өмнө
  zhch158_admin 5e7ec5c568 feat: enhance EnhancedDocPipeline with detailed element categorization, improved processing methods for text, tables, images, equations, and code, and added support for handling discarded elements and OCR box transformations 1 долоо хоног өмнө
  zhch158_admin a3f8091998 feat: introduce main_v2 script for enhanced financial document processing, supporting PDF classification, layout detection, parallel OCR and table recognition, and multi-format output options 1 долоо хоног өмнө
  zhch158_admin 7a2edafbeb feat: enhance document processing framework with updated model references, detailed layout detection strategies, and improved output formats including visualizations and coordinate information 1 долоо хоног өмнө
  zhch158_admin 63b51d4f53 feat: update utils module to include new output formatters and enhance export capabilities 1 долоо хоног өмнө
  zhch158_admin 5faab176c3 feat: add detailed output format documentation for MinerUOutputFormatter, including JSON, Markdown, and HTML specifications 1 долоо хоног өмнө
  zhch158_admin f5cee6a146 chore: remove annual report knowledge graph generation documentation 1 долоо хоног өмнө
  zhch158_admin 8182c4700d feat: add SLANet+ integration for table structure recognition and OCR result merging in Markdown documentation 1 долоо хоног өмнө
  zhch158_admin 012ee20ef6 feat: introduce enhanced output formatter v2 for document processing, supporting MinerU standard JSON, Markdown, HTML table outputs, and debug image generation 1 долоо хоног өмнө
  zhch158_admin 8ef1d66999 feat: add MinerU output formatter for standardized document processing, supporting JSON, Markdown, and table formats with image saving capabilities 1 долоо хоног өмнө
  zhch158_admin 61cca17476 refactor: remove unused cell extraction method and related logic from MinerUVLRecognizer 1 долоо хоног өмнө
  zhch158_admin f5eacc33c8 新增增强版文档处理流水线 v2,支持PDF分类、页面方向识别、并行处理文本和表格区域,以及跨页表格合并功能。 1 долоо хоног өмнө
  zhch158_admin b535f7755e 新增银行交易流水和财务报表场景配置 v2,支持多种输入格式和增强的识别功能,包括表格处理、OCR识别及跨页合并配置。 1 долоо хоног өмнө
  zhch158 487b25d669 !19 Merge pull request #4069 from opendatalab/release-2.6.5 1 долоо хоног өмнө
  Xiaomeng Zhao 50f776ad81 Merge pull request #4069 from opendatalab/release-2.6.5 1 долоо хоног өмнө
  Xiaomeng Zhao 94dcf754b3 Merge pull request #4068 from opendatalab/dev 1 долоо хоног өмнө
  Xiaomeng Zhao 77c18e958f Merge pull request #4067 from myhloli/dev 1 долоо хоног өмнө
  myhloli 16997bea1b fix: remove unnecessary dependencies from ppu.Dockerfile 1 долоо хоног өмнө
  Xiaomeng Zhao 1a15dcee32 Merge pull request #4065 from opendatalab/master 1 долоо хоног өмнө