Commit Graph

  • dcfa8bcd32 feat: update calls for default page prompt ElHachem02 2026-02-23 11:38:57 +01:00
  • ffc929bb97 feat: merge with main branch ElHachem02 2026-02-23 11:32:18 +01:00
  • 40982122af feat: define constant for base prompt and udpate codebase ElHachem02 2026-02-23 10:55:21 +01:00
  • 684f59f263 chore: add deprecation flag for annotations (#3001) Peter W. J. Staar 2026-02-19 17:18:54 +01:00
  • 03532938b5 feat: Unified model-family inference engines (including image-classification) and KServe v2 API support (#2979) Christoph Auer 2026-02-18 10:49:19 +01:00
  • 6f1f2a9ffb chore: Update outdated GitHub Actions versions (#2918) Pádraic Slattery 2026-02-18 08:03:24 +01:00
  • 460e4e5ce6 chore: bump version to 2.74.0 [skip ci] v2.74.0 github-actions[bot] 2026-02-17 21:16:42 +00:00
  • 576bada7b7 fix: security vulnerabilities with XML External Entity and related attacks (#3009) Cesar Berrospi Ramis 2026-02-17 20:29:04 +01:00
  • bf417e6d26 feat: Introduce docling-parse v5 and deprecate old docling-parse backends (#2872) Peter W. J. Staar 2026-02-17 20:27:56 +01:00
  • a1b0e3fd6b fix(csv): set default delimiter by default (#3005) Cesar Berrospi Ramis 2026-02-17 20:26:05 +01:00
  • dbba6ea27f fix: improved deserialization of engine_options (#3008) Michele Dolfi 2026-02-17 12:42:19 +01:00
  • fe3515003f chore:remove-docling-v1-and-rename-docling-v2-to-docling chore/remove-docling-v1-and-rename-docling-v2-to-docling Peter Staar 2026-02-17 06:58:24 +01:00
  • 09378ac850 updated comments for later docling-parse integrations Peter Staar 2026-02-17 05:49:48 +01:00
  • 78b989bce4 updated the groundtruth to deal with rounding errors Peter Staar 2026-02-17 05:44:30 +01:00
  • 4d0eadde13 ran pre-commit Peter Staar 2026-02-16 17:57:10 +01:00
  • 368d638f88 fixed the backend_docling_parse Peter Staar 2026-02-16 17:56:41 +01:00
  • fd5934783e ran the pre-commit Peter Staar 2026-02-16 16:09:04 +01:00
  • 54202dd301 Updated the docling-parse to 5.3.0 Peter Staar 2026-02-16 15:38:38 +01:00
  • 16b2081035 chore: bump version to 2.73.1 [skip ci] v2.73.1 github-actions[bot] 2026-02-13 15:34:52 +00:00
  • 86b691204d fix(asciidoc): handle commas in image alt text (#2983) Felix Wente 2026-02-13 15:29:02 +01:00
  • e2870f94ed fix: Use timezone-aware datetime (#2947) Nikhil Singh 2026-02-13 19:58:53 +05:30
  • 1f914826bb fix: add failed pages to DoclingDocument for page break consistency (#2939) jhchoi1182 2026-02-13 21:35:35 +09:00
  • 0967a4d908 fix:Loosen pillow version constraints to allow CVE-2026-25990 fix (#2992) Samved Divekar 2026-02-13 01:03:58 -08:00
  • 9166b47e73 chore: bump version to 2.73.0 [skip ci] v2.73.0 github-actions[bot] 2026-02-11 09:53:58 +00:00
  • 704ef0afba docs: Add LaTeX and WebVTT as supported types (#2974) Peter W. J. Staar 2026-02-10 19:59:23 +01:00
  • 14e474c955 feat: Inference engines abstraction for object detection model family with HF Transformers and ONNX runtime (#2959) Christoph Auer 2026-02-10 16:15:30 +01:00
  • e6ccb8b2c1 feat: added support for parsing LaTeX (.tex) documents (#2890) Aditya Sasidhar 2026-02-10 19:43:09 +05:30
  • 21440d81ff Fix narrow type assertions cau/allow-pipeline-options-override-without-reinit Christoph Auer 2026-02-10 13:58:03 +01:00
  • dd9eb3236a fix: enforce strict compatible pipeline overrides without reinit Christoph Auer 2026-02-10 13:34:17 +01:00
  • 34c0dbc5de Merge branch 'main' of github.com:DS4SD/docling into cau/allow-pipeline-options-override-without-reinit Christoph Auer 2026-02-10 12:21:08 +01:00
  • 9721321c46 fix: Restore expected behavior for artifacts_path and accelerator_options in VLM engines (#2961) Michele Dolfi 2026-02-09 09:46:46 +01:00
  • ae4fdbbb09 fix: allow offline chart extraction model artifacts (#2957) Michele Dolfi 2026-02-06 12:11:49 +01:00
  • 4ea1204977 Merge from main Christoph Auer 2026-02-05 15:11:19 +01:00
  • d4c87133f3 feat: Introduce pluggable VLM runtime system with preset-based configuration (#2919) Michele Dolfi 2026-02-04 17:29:17 +01:00
  • 514d99f60a Enable pipeline override and reuse with compatible options (WIP) Christoph Auer 2026-02-04 17:15:10 +01:00
  • e1e52b01af rename runtime to inference engine Michele Dolfi 2026-02-04 12:40:45 +01:00
  • e186e6cf7a Merge remote-tracking branch 'origin/main' into feat-model-runtimes Michele Dolfi 2026-02-04 10:27:10 +01:00
  • 256d9a2249 update docs catalog page Michele Dolfi 2026-02-03 19:02:01 +01:00
  • 92a7e8d3d0 add docs with stages Michele Dolfi 2026-02-03 17:16:09 +01:00
  • c28d0d6c5d chore: bump version to 2.72.0 [skip ci] v2.72.0 github-actions[bot] 2026-02-03 15:08:32 +00:00
  • 356bfa0198 fix test Michele Dolfi 2026-02-03 15:05:15 +01:00
  • bbf4821481 fixes Michele Dolfi 2026-02-03 14:40:17 +01:00
  • 2259a55cfe renaming from runtime to inference engine and model families Michele Dolfi 2026-02-03 14:30:01 +01:00
  • a5ad8f24ff docs: add granite vision for charts (#2946) Peter W. J. Staar 2026-02-03 12:31:27 +01:00
  • fe45c71fe7 feat: add chart extraction models (#2848) Peter W. J. Staar 2026-02-03 09:03:41 +01:00
  • c2edf64a16 rename runtimes to explicit vlm_runtimes Michele Dolfi 2026-02-02 10:33:30 +01:00
  • 3110c439da fix(backend): improve Excel table bounds detection and flatten merged cells (#2778) Rashid Ul Islam 2026-02-02 15:01:11 +05:30
  • 474d00ec0f flasg for CI Michele Dolfi 2026-02-02 09:56:43 +01:00
  • 053e611761 use new vlm runtime class Michele Dolfi 2026-02-02 09:55:11 +01:00
  • c07c3b1028 move vlm_convert_model Michele Dolfi 2026-02-01 22:05:24 +01:00
  • e65bd75465 avoid automatic fallback to mlx and fix end_of_utterance in codeformula Michele Dolfi 2026-02-01 22:00:19 +01:00
  • 8dc0fcd232 fix test Michele Dolfi 2026-02-01 21:14:49 +01:00
  • 1c0b53a24e add another legacy example Michele Dolfi 2026-02-01 21:10:44 +01:00
  • 036b659a8d fix legacy examples Michele Dolfi 2026-02-01 21:10:17 +01:00
  • 7b96837f15 update vlm api model example Michele Dolfi 2026-02-01 20:46:49 +01:00
  • eb6c53ad7e Merge remote-tracking branch 'origin/main' into feat-model-runtimes Michele Dolfi 2026-02-01 20:15:01 +01:00
  • ab748a2b35 remove unused repo_id Michele Dolfi 2026-02-01 19:12:36 +01:00
  • afa2d3664c add all models to presets and run compare_vlm Michele Dolfi 2026-02-01 18:59:05 +01:00
  • 1a3d2b0bf3 fix running minimal_vlm example Michele Dolfi 2026-02-01 17:15:49 +01:00
  • daa90bf262 rename code formula presets Michele Dolfi 2026-02-01 17:08:01 +01:00
  • 334ae81bcf add granite docling as code formula model Michele Dolfi 2026-02-01 16:57:49 +01:00
  • 76f986b856 working picture description examples Michele Dolfi 2026-02-01 16:35:05 +01:00
  • aa0bb26b20 remove duplicated predict() and factor out some utils Michele Dolfi 2026-02-01 15:46:53 +01:00
  • 5e452a2e8f fix(pptx): handle picture shapes with external image references (#2914) Sam Quigley 2026-02-01 05:44:29 -05:00
  • 6278eb5b0e use chat template Michele Dolfi 2026-01-30 19:43:45 +01:00
  • 1d6264cf33 per stage registry Michele Dolfi 2026-01-30 19:26:41 +01:00
  • 7957842825 update all stages with original setup Michele Dolfi 2026-01-30 19:24:53 +01:00
  • 1cfbcfdf27 revisit init logic and propagate the proper options to the runtimes Michele Dolfi 2026-01-30 19:01:03 +01:00
  • c74d378b08 chore: bump version to 2.71.0 [skip ci] v2.71.0 github-actions[bot] 2026-01-30 17:11:20 +00:00
  • 46188c1a62 use granite 3.3 and set options Michele Dolfi 2026-01-30 17:48:36 +01:00
  • 0602a7cdab feat: webvtt and source tracker (#2787) Cesar Berrospi Ramis 2026-01-30 17:44:03 +01:00
  • dc406cd10f update model Michele Dolfi 2026-01-30 13:48:09 +01:00
  • 0e1007ad0e keep old stage Michele Dolfi 2026-01-30 13:28:34 +01:00
  • f48d8b4c8c fixes for running examples Michele Dolfi 2026-01-30 10:28:31 +01:00
  • dfb610e1ea update examples Michele Dolfi 2026-01-29 18:05:22 +01:00
  • daedeeecdc running Michele Dolfi 2026-01-28 20:18:58 +01:00
  • f9b803e71a use new model settings by default Michele Dolfi 2026-01-28 17:41:53 +01:00
  • 35da1f8fa4 use presets and new vlm options in CLI Michele Dolfi 2026-01-28 14:51:07 +01:00
  • 6f205ae211 fix: allow newer typer versions (#2930) Michele Dolfi 2026-01-28 14:42:38 +01:00
  • 82b7982e1b fix(rapidocr): Use new model links for RapidOCR (#2928) nkh0472 2026-01-28 18:44:27 +08:00
  • 4a269de91a fix: presets for ollama (#2926) Michele Dolfi 2026-01-28 11:44:13 +01:00
  • ab29cee181 batch prediction Michele Dolfi 2026-01-27 08:29:46 +01:00
  • a8cae1eadd fix code formula preset Michele Dolfi 2026-01-27 08:18:55 +01:00
  • f739c1e81f chore: Add CI tests for pip install without lock file (#2909) Michele Dolfi 2026-01-26 17:00:36 +01:00
  • d5b7e2df08 add test Michele Dolfi 2026-01-26 13:45:19 +01:00
  • d355a79e2a Merge remote-tracking branch 'origin/main' into feat-model-runtimes Michele Dolfi 2026-01-26 13:27:48 +01:00
  • 9385731f79 model runtime refactoring Michele Dolfi 2026-01-26 13:26:40 +01:00
  • c49d7f36a6 merged with main Peter Staar 2026-01-26 11:31:15 +01:00
  • b6ca094519 feat: add support for Word document comments extraction (#2834) Siva 2026-01-26 14:28:46 +05:30
  • e413e688ed chore: bump version to 2.70.0 [skip ci] v2.70.0 github-actions[bot] 2026-01-23 15:22:52 +00:00
  • 86eaef5b45 fix(md): handle pipe symbols that are not table markers (#2904) Cesar Berrospi Ramis 2026-01-23 15:19:09 +01:00
  • 7f386587ed feat: Drop support for Python 3.9 (#2905) Michele Dolfi 2026-01-23 10:15:58 +01:00
  • 7a1952ae3d fix: remove direct vllm dependency (#2910) Michele Dolfi 2026-01-22 19:41:21 +01:00
  • 412e23e29a chore: upgrade locked deps (#2906) Michele Dolfi 2026-01-22 17:22:13 +01:00
  • 002096bd3c merged with main Peter Staar 2026-01-22 10:46:27 -05:00
  • c1dd01c6c1 added the test_backend_docling_parse Peter Staar 2026-01-22 10:44:36 -05:00
  • ab91786f3b docs: add comprehensive docstrings to PdfPipelineOptions (#2827) Nalini Panwar 2026-01-22 15:29:11 +05:30
  • 999dbb2765 fix: PPTX parsing: bullet points not grouped correctly under subheadings (#2663) (#2855) Tong Luo 2026-01-21 21:34:24 +08:00
  • cf934eaa20 chore: bump version to 2.69.1 [skip ci] v2.69.1 github-actions[bot] 2026-01-21 12:45:50 +00:00
  • 08f49e2abc fix: off-by-one error for page indexing in vlm_pipeline (#2902) Christoph Auer 2026-01-21 13:11:52 +01:00