Commit Graph

  • a2e8b81dd1 chore: Code comments cau/kserve-ocr-model Nikos Livathinos 2026-03-26 09:54:27 +01:00
  • f2b716d4cd Revert debug logging Christoph Auer 2026-03-25 16:48:18 +01:00
  • c38534ddb3 Reduce scale to 2.0 Christoph Auer 2026-03-25 16:15:24 +01:00
  • eeb16407c1 Add debug logging Christoph Auer 2026-03-25 15:38:49 +01:00
  • aaac9879b6 Add BYTES support to KServe v2 client Christoph Auer 2026-03-25 15:33:00 +01:00
  • 90af2cbcd4 Modify batch dimension handling again Christoph Auer 2026-03-25 15:11:12 +01:00
  • 34f86ddcf8 Modify batch dimension handling Christoph Auer 2026-03-25 13:58:05 +01:00
  • 6532ec66b8 Add kserve_v2_ocr as default in factory Christoph Auer 2026-03-25 13:53:52 +01:00
  • dccdd68f7d feat: add KServe v2 OCR engine with mixin refactoring Christoph Auer 2026-03-25 13:13:54 +01:00
  • 1a9a29cc8e Deployed 53412ed with MkDocs version: 1.6.1 gh-pages 2026-03-25 09:46:09 +00:00
  • 53412ed43c chore: re-export DocumentStream explicitly for type checkers (#3181) main Eugenio-BAYE 2026-03-25 10:43:01 +01:00
  • d355af6c59 chore: bump version to 2.82.0 [skip ci] v2.82.0 github-actions[bot] 2026-03-25 09:40:06 +00:00
  • e36125ba2d fix(omml): correct LaTeX output for fractions, math operators, and functions (#3122) Giulio Leone 2026-03-25 07:07:31 +01:00
  • a0fc3c9d73 fix: manage PDFium backend resource lifecycles to avoid SIGSEGV/SIGTRAP crashes (#3180) Christoph Auer 2026-03-24 17:59:55 +01:00
  • 1c74a9b9c7 feat: Implementation of HTML backend with headless browser (#2969) Maxim Lysak 2026-03-24 14:28:57 +01:00
  • 79b18945a8 Merge branch 'main' of github.com:DS4SD/docling into cau/pdfium-bitmap-fix cau/pdfium-bitmap-fix Christoph Auer 2026-03-24 13:15:32 +01:00
  • 70bf513ea7 upgrade uv.lock Christoph Auer 2026-03-24 13:15:22 +01:00
  • 36c219551e upgrade uv.lock cau/pdfium-race-fix Christoph Auer 2026-03-24 13:02:51 +01:00
  • b3f4e6692d fix: explicitly close PdfBitmap after copy in both PDF backends Christoph Auer 2026-03-24 12:52:48 +01:00
  • 5c63b94009 fix: explicitly close PdfBitmap after copy in both PDF backends Christoph Auer 2026-03-24 10:34:39 +01:00
  • 90d6dd4e87 fix(docx): split multiple OMML equations into separate formula items (#3123) Giulio Leone 2026-03-24 09:42:16 +01:00
  • bf3b6cad34 Add missing file Christoph Auer 2026-03-23 17:24:13 +01:00
  • b56a0195f1 fix: harden pdfium backend lifecycle and threaded shutdown Christoph Auer 2026-03-23 17:09:41 +01:00
  • d520eb0ece Switch to the latest version of DocumentFigureClassifier model v2.5 vku/docfm_v25 Viktor Kuropiatnyk 2026-03-23 16:05:00 +01:00
  • 63d6425d4a Merge branch 'main' of github.com:DS4SD/docling into cau/pdfium-race-fix Christoph Auer 2026-03-23 16:01:00 +01:00
  • fdf5e20ccd fix: let user params override engine defaults in API VLM engine (#3116) lif 2026-03-23 18:38:00 +08:00
  • f0e3d1df2a fix(vlm): handle content_filter finish reason in API responses (#3051) Br1an 2026-03-23 17:46:08 +08:00
  • fffd445789 docs: Fix Erroneous vLLM VLM pipeline engine option params causing empty/bad responses (#3167) Max Swain 2026-03-23 09:44:37 +00:00
  • 5473e07450 fix(cli): avoid generating images for non-image exports (#3127) Muhammad Hassan Raza 2026-03-23 12:16:32 +05:00
  • 9abf0fd385 fix: honor picture description batching and scale options (#3132) Muhammad Hassan Raza 2026-03-23 12:14:51 +05:00
  • 4e650af56d chore: bump version to 2.81.0 [skip ci] v2.81.0 github-actions[bot] 2026-03-20 21:32:59 +00:00
  • 2f7c09e0d8 fix(docx): Missing list items after numbered header (#2665) (#2678) Emre Çalışır 2026-03-20 23:24:27 +03:00
  • 96d7c7ec79 feat: route plain-text and Quarto/R Markdown files to the Markdown backend (#3161) Peter W. J. Staar 2026-03-20 16:38:16 +01:00
  • afb4bb6802 fix: Avoid thread-unsafe close of pypdfium backend (#3160) Christoph Auer 2026-03-20 12:43:22 +01:00
  • fca13cf18f fix: Avoid thread-unsafe close of pypdfium backend Christoph Auer 2026-03-20 11:43:55 +01:00
  • 8ae0974a9d fix: handle external image relationships in MsWordDocumentBackend (#3114) Ron 2026-03-19 15:22:21 +02:00
  • a39317a147 fix: handle PermissionError for directory input on Windows CLI (#3149) Joaquin Hui 2026-03-19 13:06:17 +00:00
  • b6a37b789c change to sentence-level OCR cau/nvidia-nemotron-ocr Christoph Auer 2026-03-17 14:51:11 +01:00
  • 412af62135 fix: avoid in-place mutation of pipeline options breaking cache key (#3115) lif 2026-03-17 16:59:38 +08:00
  • 53a5f80a43 fix: preserve torch_dtype in get_engine_config and add it to CodeFormulaV2 (#3117) lif 2026-03-17 16:59:09 +08:00
  • 0447a92ff2 change to --no-extra Christoph Auer 2026-03-16 21:47:59 +01:00
  • d0d0f69ec6 Remove the monkeypatch Christoph Auer 2026-03-16 18:29:45 +01:00
  • b200c65eb1 Pin exactly torch 2.9.1 for nemotron-ocr extra Christoph Auer 2026-03-16 18:22:39 +01:00
  • 664bd7d92a docs. update documentation Christoph Auer 2026-03-16 15:04:36 +01:00
  • edb620b1f4 Make minimal monkey-patch Christoph Auer 2026-03-16 15:00:55 +01:00
  • fc2ab41b50 Add model downloader support Christoph Auer 2026-03-16 14:53:11 +01:00
  • fa6e5fb13c debug: roll back debug printing Christoph Auer 2026-03-16 14:31:14 +01:00
  • 9de1985771 debug: test grid sampler monkeypatch Christoph Auer 2026-03-16 14:16:00 +01:00
  • 52bebae429 debug: Add prints to find out where predictions fail Christoph Auer 2026-03-16 13:49:40 +01:00
  • 1e841ebcbd fix: release image backend resources after frame extraction (#3134) Abi 2026-03-16 18:00:22 +05:30
  • f8cf4b0025 feat: Add NVIDIA nemotron-ocr as supporetd backend Christoph Auer 2026-03-16 13:16:02 +01:00
  • ce4992363f test: fix picture checks in shared verifier (#3130) Muhammad Hassan Raza 2026-03-16 15:43:43 +05:00
  • 6238aa35d0 chore: address stale TODO comments across codebase (#3108) Muhammad Hassan Raza 2026-03-14 18:01:25 +05:00
  • cbe3db470d chore: bump version to 2.80.0 [skip ci] v2.80.0 github-actions[bot] 2026-03-14 05:57:47 +00:00
  • f950679f60 feat: add the VllmCudaGraphMode (#3125) Peter W. J. Staar 2026-03-13 15:23:25 +01:00
  • f73df4f916 chore: bump version to 2.79.0 [skip ci] v2.79.0 github-actions[bot] 2026-03-12 07:40:51 +00:00
  • 93f6feeabc fix: Use OCR cells with TableFormer v2 (#3107) Christoph Auer 2026-03-11 21:51:26 +01:00
  • 2a0e11f762 fix: add self-consistency check in the table-structure model (#3105) Peter W. J. Staar 2026-03-11 17:13:53 +01:00
  • 7952efee2f feat: add fact metadata and linkbase relationships for XBRL (#3084) Darshan 2026-03-11 11:28:35 -04:00
  • 198d0af19b fix: correct typos in log messages and add missing error log (#3097) Muhammad Hassan Raza 2026-03-11 20:08:03 +05:00
  • 4b66f1453a refactor(backend): modularize latex backend into a package structure (#3092) Aditya Sasidhar 2026-03-11 20:37:30 +05:30
  • fb96f288f2 test: fix missing assert in page_count tests (#3102) Muhammad Hassan Raza 2026-03-11 18:29:49 +05:00
  • fef01f8c88 fix: Don't force cast to float32 in API Kserve v2 inputs (#3101) Christoph Auer 2026-03-10 18:45:57 +01:00
  • 594fc3e2ca chore: bump version to 2.78.0 [skip ci] v2.78.0 github-actions[bot] 2026-03-10 14:55:08 +00:00
  • 4ccd1d465d feat: Add support for TableFormer v2 (#3013) Peter W. J. Staar 2026-03-10 11:57:00 +01:00
  • 80f75b8896 fix(html): fix broken document tree and quadratic complexity in rich table cells (#3025) Ivan Traus 2026-03-10 01:48:21 -07:00
  • d113e611c4 docs: fix code in rag langchain chunker tokenizer (#2993) Maree Carroll 2026-03-10 16:50:24 +11:00
  • 5188180ea3 fix: loosen dependency for pandas3 (#3095) Faiq Adzlan 2026-03-10 13:36:19 +08:00
  • 3b7bba0212 chore: Revert unintended test ground truth changes from #3019 (#3093) Christoph Auer 2026-03-09 17:38:34 +01:00
  • 1192714b53 fix: add parse timeout to legacy LaTeX documents (#3019) Aditya Sasidhar 2026-03-09 15:22:56 +05:30
  • ee16285651 fix(msword): skip GroupItem targets without comments attribute (#3080) JiangNan 2026-03-09 17:08:27 +08:00
  • 95b759e519 docs: update code snippet to use modern pipeline options syntax (#3087) Robert Sokolewicz 2026-03-09 08:49:53 +01:00
  • 5d3ac38a65 docs: set HuggingFaceEndpoint task for Mixtral examples (#2945) Kaiiiiiiiii 2026-03-08 17:38:06 +08:00
  • 3d90778e3e feat: Add gRPC transport for KServe v2 API engine (#3074) Christoph Auer 2026-03-07 10:54:25 +01:00
  • b7815658d1 chore: bump version to 2.77.0 [skip ci] v2.77.0 github-actions[bot] 2026-03-06 13:45:28 +00:00
  • cd9dd10ccf fix(docx): preserve URL fragments and query params in hyperlinks (#3050) Br1an 2026-03-06 18:35:17 +08:00
  • 38c4bb26e8 feat: track vlm_inference time for mlx_model pipeline (#3060) Peter El Hachem 2026-03-06 11:26:45 +01:00
  • 56f06fe372 fix: detect Office Open XML formats from ZIP contents when filename has no extension (#3073) Debaditya Shome 2026-03-06 05:01:30 -05:00
  • cfc6636a2a feat: add configurable graph_optimization_level for ONNX Runtime engines (#3071) Aditya Sasidhar 2026-03-06 15:13:12 +05:30
  • 653940e025 docs: add examples and fix docstring bug in DocumentConverter (#3064) easedu 2026-03-06 06:41:37 -03:00
  • 8b990856cd docs: add docstrings to PipelineOptions classes (#3065) easedu 2026-03-06 06:16:13 -03:00
  • f7cb304daa fix(readingorder): assign FURNITURE content_layer to footer/header in container groups (#3044) Giulio Leone 2026-03-06 09:37:16 +01:00
  • 56eb12782c fix(docx): handle list items immediately after numbered headings (#3070) Cesar Berrospi Ramis 2026-03-06 09:30:48 +01:00
  • 68336c2bda fix(rapidocr): ORT thread configuration for RapidOCR backend (#3062) BODAPATIMAHESH 2026-03-06 13:41:55 +05:30
  • 752f81b3dd chore: bump version to 2.76.0 [skip ci] v2.76.0 github-actions[bot] 2026-03-02 14:43:12 +00:00
  • c90c068600 feat: add documentation on why we set the layout model to run on CPU elh/update_2stage_inference ElHachem02 2026-03-02 15:16:46 +01:00
  • 859c302310 fix(xlsx): handle OneCellAnchor images in Excel backend (#3045) Br1an 2026-03-02 19:55:17 +08:00
  • 6198e69dec fix: normalize Unicode ligatures in PDF text extraction (#3057) Michele Dolfi 2026-03-02 11:11:15 +01:00
  • 7aacc6c18d docs: add metaxy integration (#3058) geoHeil 2026-03-02 09:48:35 +01:00
  • 672125cd1b docs: removes merge conflict artifacts (#3055) Phil Nash 2026-03-02 18:17:02 +11:00
  • 477359b772 fix(ocr): update RapidOCR torch GPU config key (#3049) Br1an 2026-03-02 14:15:20 +08:00
  • 90ce93d8a0 fix: convert PIL images to RGB before picture description (#3014) Aatrey Sahay 2026-03-01 06:20:41 -05:00
  • 1321b39cd8 docs: add audio & video processing guide (#3038) Tejas Kumar 2026-03-01 09:00:48 +01:00
  • 1eb5c21dab docs: add XBRL conversion example notebook and update feature listings (#3039) Cesar Berrospi Ramis 2026-02-27 16:09:19 +01:00
  • d276e60561 feat: export to WebVTT format (#3036) Cesar Berrospi Ramis 2026-02-27 14:22:52 +01:00
  • a3d2b4bcc0 fix(msword): use outlineLvl for heading levels and clamp to minimum 1 (#2916) Sam Quigley 2026-02-25 11:02:36 -05:00
  • aec57629af chore: bump version to 2.75.0 [skip ci] v2.75.0 github-actions[bot] 2026-02-24 20:16:56 +00:00
  • 334ba6e51f feat: create a backend parser for XBRL instance reports (#3017) Cesar Berrospi Ramis 2026-02-24 16:52:02 +01:00
  • 6b824f844b fix: skip ASR segments when length is zero (#2998) Scott McKendry 2026-02-24 19:18:41 +13:00
  • 236216ed4e fix(docx): guard against None hyperlink address in _get_paragraph_elements (#2367) (#3022) Hemant Sudarshan 2026-02-24 10:26:45 +05:30