docling 2.34.0__py3-none-any.whl → 2.36.0__py3-none-any.whl

docling/pipeline/vlm_pipeline.py

@@ -1,29 +1,46 @@
 import logging
+import re
 from io import BytesIO
 from pathlib import Path
 from typing import List, Optional, Union, cast
 
-from docling_core.types import DoclingDocument
-from docling_core.types.doc import BoundingBox, DocItem, ImageRef, PictureItem, TextItem
+from docling_core.types.doc import (
+    BoundingBox,
+    DocItem,
+    DoclingDocument,
+    ImageRef,
+    PictureItem,
+    ProvenanceItem,
+    TextItem,
+)
+from docling_core.types.doc.base import (
+    BoundingBox,
+    Size,
+)
 from docling_core.types.doc.document import DocTagsDocument
 from PIL import Image as PILImage
 
 from docling.backend.abstract_backend import AbstractDocumentBackend
+from docling.backend.html_backend import HTMLDocumentBackend
 from docling.backend.md_backend import MarkdownDocumentBackend
 from docling.backend.pdf_backend import PdfDocumentBackend
 from docling.datamodel.base_models import InputFormat, Page
 from docling.datamodel.document import ConversionResult, InputDocument
 from docling.datamodel.pipeline_options import (
+    VlmPipelineOptions,
+)
+from docling.datamodel.pipeline_options_vlm_model import (
     ApiVlmOptions,
-    HuggingFaceVlmOptions,
     InferenceFramework,
+    InlineVlmOptions,
     ResponseFormat,
-    VlmPipelineOptions,
 )
 from docling.datamodel.settings import settings
 from docling.models.api_vlm_model import ApiVlmModel
-from docling.models.hf_mlx_model import HuggingFaceMlxModel
-from docling.models.hf_vlm_model import HuggingFaceVlmModel
+from docling.models.vlm_models_inline.hf_transformers_model import (
+    HuggingFaceTransformersVlmModel,
+)
+from docling.models.vlm_models_inline.mlx_model import HuggingFaceMlxModel
 from docling.pipeline.base_pipeline import PaginatedPipeline
 from docling.utils.profiling import ProfilingScope, TimeRecorder
 
@@ -66,8 +83,8 @@ class VlmPipeline(PaginatedPipeline):
                     vlm_options=cast(ApiVlmOptions, self.pipeline_options.vlm_options),
                 ),
             ]
-        elif isinstance(self.pipeline_options.vlm_options, HuggingFaceVlmOptions):
-            vlm_options = cast(HuggingFaceVlmOptions, self.pipeline_options.vlm_options)
+        elif isinstance(self.pipeline_options.vlm_options, InlineVlmOptions):
+            vlm_options = cast(InlineVlmOptions, self.pipeline_options.vlm_options)
             if vlm_options.inference_framework == InferenceFramework.MLX:
                 self.build_pipe = [
                     HuggingFaceMlxModel(
@@ -77,15 +94,19 @@ class VlmPipeline(PaginatedPipeline):
                         vlm_options=vlm_options,
                     ),
                 ]
-            else:
+            elif vlm_options.inference_framework == InferenceFramework.TRANSFORMERS:
                 self.build_pipe = [
-                    HuggingFaceVlmModel(
+                    HuggingFaceTransformersVlmModel(
                         enabled=True,  # must be always enabled for this pipeline to make sense.
                         artifacts_path=artifacts_path,
                         accelerator_options=pipeline_options.accelerator_options,
                         vlm_options=vlm_options,
                     ),
                 ]
+            else:
+                raise ValueError(
+                    f"Could not instantiate the right type of VLM pipeline: {vlm_options.inference_framework}"
+                )
 
         self.enrichment_pipe = [
             # Other models working on `NodeItem` elements in the DoclingDocument
@@ -116,49 +137,19 @@ class VlmPipeline(PaginatedPipeline):
                 self.pipeline_options.vlm_options.response_format
                 == ResponseFormat.DOCTAGS
             ):
-                doctags_list = []
-                image_list = []
-                for page in conv_res.pages:
-                    predicted_doctags = ""
-                    img = PILImage.new("RGB", (1, 1), "rgb(255,255,255)")
-                    if page.predictions.vlm_response:
-                        predicted_doctags = page.predictions.vlm_response.text
-                    if page.image:
-                        img = page.image
-                    image_list.append(img)
-                    doctags_list.append(predicted_doctags)
-
-                doctags_list_c = cast(List[Union[Path, str]], doctags_list)
-                image_list_c = cast(List[Union[Path, PILImage.Image]], image_list)
-                doctags_doc = DocTagsDocument.from_doctags_and_image_pairs(
-                    doctags_list_c, image_list_c
-                )
-                conv_res.document = DoclingDocument.load_from_doctags(doctags_doc)
-
-                # If forced backend text, replace model predicted text with backend one
-                if self.force_backend_text:
-                    scale = self.pipeline_options.images_scale
-                    for element, _level in conv_res.document.iterate_items():
-                        if not isinstance(element, TextItem) or len(element.prov) == 0:
-                            continue
-                        page_ix = element.prov[0].page_no - 1
-                        page = conv_res.pages[page_ix]
-                        if not page.size:
-                            continue
-                        crop_bbox = (
-                            element.prov[0]
-                            .bbox.scaled(scale=scale)
-                            .to_top_left_origin(page_height=page.size.height * scale)
-                        )
-                        txt = self.extract_text_from_backend(page, crop_bbox)
-                        element.text = txt
-                        element.orig = txt
+                conv_res.document = self._turn_dt_into_doc(conv_res)
+
             elif (
                 self.pipeline_options.vlm_options.response_format
                 == ResponseFormat.MARKDOWN
             ):
                 conv_res.document = self._turn_md_into_doc(conv_res)
 
+            elif (
+                self.pipeline_options.vlm_options.response_format == ResponseFormat.HTML
+            ):
+                conv_res.document = self._turn_html_into_doc(conv_res)
+
             else:
                 raise RuntimeError(
                     f"Unsupported VLM response format {self.pipeline_options.vlm_options.response_format}"
@@ -192,23 +183,199 @@ class VlmPipeline(PaginatedPipeline):
 
         return conv_res
 
-    def _turn_md_into_doc(self, conv_res):
-        predicted_text = ""
-        for pg_idx, page in enumerate(conv_res.pages):
+    def _turn_dt_into_doc(self, conv_res) -> DoclingDocument:
+        doctags_list = []
+        image_list = []
+        for page in conv_res.pages:
+            predicted_doctags = ""
+            img = PILImage.new("RGB", (1, 1), "rgb(255,255,255)")
             if page.predictions.vlm_response:
-                predicted_text += page.predictions.vlm_response.text + "\n\n"
-        response_bytes = BytesIO(predicted_text.encode("utf8"))
-        out_doc = InputDocument(
-            path_or_stream=response_bytes,
-            filename=conv_res.input.file.name,
-            format=InputFormat.MD,
-            backend=MarkdownDocumentBackend,
+                predicted_doctags = page.predictions.vlm_response.text
+            if page.image:
+                img = page.image
+            image_list.append(img)
+            doctags_list.append(predicted_doctags)
+
+        doctags_list_c = cast(List[Union[Path, str]], doctags_list)
+        image_list_c = cast(List[Union[Path, PILImage.Image]], image_list)
+        doctags_doc = DocTagsDocument.from_doctags_and_image_pairs(
+            doctags_list_c, image_list_c
         )
-        backend = MarkdownDocumentBackend(
-            in_doc=out_doc,
-            path_or_stream=response_bytes,
+        conv_res.document = DoclingDocument.load_from_doctags(
+            doctag_document=doctags_doc
         )
-        return backend.convert()
+
+        # If forced backend text, replace model predicted text with backend one
+        if page.size:
+            if self.force_backend_text:
+                scale = self.pipeline_options.images_scale
+                for element, _level in conv_res.document.iterate_items():
+                    if not isinstance(element, TextItem) or len(element.prov) == 0:
+                        continue
+                    crop_bbox = (
+                        element.prov[0]
+                        .bbox.scaled(scale=scale)
+                        .to_top_left_origin(page_height=page.size.height * scale)
+                    )
+                    txt = self.extract_text_from_backend(page, crop_bbox)
+                    element.text = txt
+                    element.orig = txt
+
+        return conv_res.document
+
+    def _turn_md_into_doc(self, conv_res):
+        def _extract_markdown_code(text):
+            """
+            Extracts text from markdown code blocks (enclosed in triple backticks).
+            If no code blocks are found, returns the original text.
+
+            Args:
+                text (str): Input text that may contain markdown code blocks
+
+            Returns:
+                str: Extracted code if code blocks exist, otherwise original text
+            """
+            # Regex pattern to match content between triple backticks
+            # This handles multiline content and optional language specifier
+            pattern = r"^```(?:\w*\n)?(.*?)```(\n)*$"
+
+            # Search with DOTALL flag to match across multiple lines
+            mtch = re.search(pattern, text, re.DOTALL)
+
+            if mtch:
+                # Return only the content of the first capturing group
+                return mtch.group(1)
+            else:
+                # No code blocks found, return original text
+                return text
+
+        for pg_idx, page in enumerate(conv_res.pages):
+            page_no = pg_idx + 1  # FIXME: might be incorrect
+
+            predicted_text = ""
+            if page.predictions.vlm_response:
+                predicted_text = page.predictions.vlm_response.text + "\n\n"
+
+            predicted_text = _extract_markdown_code(text=predicted_text)
+
+            response_bytes = BytesIO(predicted_text.encode("utf8"))
+            out_doc = InputDocument(
+                path_or_stream=response_bytes,
+                filename=conv_res.input.file.name,
+                format=InputFormat.MD,
+                backend=MarkdownDocumentBackend,
+            )
+            backend = MarkdownDocumentBackend(
+                in_doc=out_doc,
+                path_or_stream=response_bytes,
+            )
+            page_doc = backend.convert()
+
+            if page.image is not None:
+                pg_width = page.image.width
+                pg_height = page.image.height
+            else:
+                pg_width = 1
+                pg_height = 1
+
+            conv_res.document.add_page(
+                page_no=page_no,
+                size=Size(width=pg_width, height=pg_height),
+                image=ImageRef.from_pil(image=page.image, dpi=72)
+                if page.image
+                else None,
+            )
+
+            for item, level in page_doc.iterate_items():
+                item.prov = [
+                    ProvenanceItem(
+                        page_no=pg_idx + 1,
+                        bbox=BoundingBox(
+                            t=0.0, b=0.0, l=0.0, r=0.0
+                        ),  # FIXME: would be nice not to have to "fake" it
+                        charspan=[0, 0],
+                    )
+                ]
+                conv_res.document.append_child_item(child=item)
+
+        return conv_res.document
+
+    def _turn_html_into_doc(self, conv_res):
+        def _extract_html_code(text):
+            """
+            Extracts text from markdown code blocks (enclosed in triple backticks).
+            If no code blocks are found, returns the original text.
+
+            Args:
+                text (str): Input text that may contain markdown code blocks
+
+            Returns:
+                str: Extracted code if code blocks exist, otherwise original text
+            """
+            # Regex pattern to match content between triple backticks
+            # This handles multiline content and optional language specifier
+            pattern = r"^```(?:\w*\n)?(.*?)```(\n)*$"
+
+            # Search with DOTALL flag to match across multiple lines
+            mtch = re.search(pattern, text, re.DOTALL)
+
+            if mtch:
+                # Return only the content of the first capturing group
+                return mtch.group(1)
+            else:
+                # No code blocks found, return original text
+                return text
+
+        for pg_idx, page in enumerate(conv_res.pages):
+            page_no = pg_idx + 1  # FIXME: might be incorrect
+
+            predicted_text = ""
+            if page.predictions.vlm_response:
+                predicted_text = page.predictions.vlm_response.text + "\n\n"
+
+            predicted_text = _extract_html_code(text=predicted_text)
+
+            response_bytes = BytesIO(predicted_text.encode("utf8"))
+            out_doc = InputDocument(
+                path_or_stream=response_bytes,
+                filename=conv_res.input.file.name,
+                format=InputFormat.MD,
+                backend=HTMLDocumentBackend,
+            )
+            backend = HTMLDocumentBackend(
+                in_doc=out_doc,
+                path_or_stream=response_bytes,
+            )
+            page_doc = backend.convert()
+
+            if page.image is not None:
+                pg_width = page.image.width
+                pg_height = page.image.height
+            else:
+                pg_width = 1
+                pg_height = 1
+
+            conv_res.document.add_page(
+                page_no=page_no,
+                size=Size(width=pg_width, height=pg_height),
+                image=ImageRef.from_pil(image=page.image, dpi=72)
+                if page.image
+                else None,
+            )
+
+            for item, level in page_doc.iterate_items():
+                item.prov = [
+                    ProvenanceItem(
+                        page_no=pg_idx + 1,
+                        bbox=BoundingBox(
+                            t=0.0, b=0.0, l=0.0, r=0.0
+                        ),  # FIXME: would be nice not to have to "fake" it
+                        charspan=[0, 0],
+                    )
+                ]
+                conv_res.document.append_child_item(child=item)
+
+        return conv_res.document
 
     @classmethod
     def get_default_options(cls) -> VlmPipelineOptions:

docling/utils/accelerator_utils.py

@@ -1,13 +1,16 @@
 import logging
+from typing import List, Optional
 
 import torch
 
-from docling.datamodel.pipeline_options import AcceleratorDevice
+from docling.datamodel.accelerator_options import AcceleratorDevice
 
 _log = logging.getLogger(__name__)
 
 
-def decide_device(accelerator_device: str) -> str:
+def decide_device(
+    accelerator_device: str, supported_devices: Optional[List[AcceleratorDevice]] = None
+) -> str:
     r"""
     Resolve the device based on the acceleration options and the available devices in the system.
 
@@ -20,6 +23,18 @@ def decide_device(accelerator_device: str) -> str:
     has_cuda = torch.backends.cuda.is_built() and torch.cuda.is_available()
     has_mps = torch.backends.mps.is_built() and torch.backends.mps.is_available()
 
+    if supported_devices is not None:
+        if has_cuda and AcceleratorDevice.CUDA not in supported_devices:
+            _log.info(
+                f"Removing CUDA from available devices because it is not in {supported_devices=}"
+            )
+            has_cuda = False
+        if has_mps and AcceleratorDevice.MPS not in supported_devices:
+            _log.info(
+                f"Removing MPS from available devices because it is not in {supported_devices=}"
+            )
+            has_mps = False
+
     if accelerator_device == AcceleratorDevice.AUTO.value:  # Handle 'auto'
         if has_cuda:
             device = "cuda:0"

docling/utils/model_downloader.py

@@ -4,18 +4,20 @@ from typing import Optional
 
 from docling.datamodel.pipeline_options import (
     granite_picture_description,
-    smoldocling_vlm_conversion_options,
-    smoldocling_vlm_mlx_conversion_options,
     smolvlm_picture_description,
 )
 from docling.datamodel.settings import settings
+from docling.datamodel.vlm_model_specs import (
+    SMOLDOCLING_MLX,
+    SMOLDOCLING_TRANSFORMERS,
+)
 from docling.models.code_formula_model import CodeFormulaModel
 from docling.models.document_picture_classifier import DocumentPictureClassifier
 from docling.models.easyocr_model import EasyOcrModel
-from docling.models.hf_vlm_model import HuggingFaceVlmModel
 from docling.models.layout_model import LayoutModel
 from docling.models.picture_description_vlm_model import PictureDescriptionVlmModel
 from docling.models.table_structure_model import TableStructureModel
+from docling.models.utils.hf_model_download import download_hf_model
 
 _log = logging.getLogger(__name__)
 
@@ -75,7 +77,7 @@ def download_models(
 
     if with_smolvlm:
         _log.info("Downloading SmolVlm model...")
-        PictureDescriptionVlmModel.download_models(
+        download_hf_model(
             repo_id=smolvlm_picture_description.repo_id,
             local_dir=output_dir / smolvlm_picture_description.repo_cache_folder,
             force=force,
@@ -84,26 +86,25 @@ def download_models(
 
     if with_smoldocling:
         _log.info("Downloading SmolDocling model...")
-        HuggingFaceVlmModel.download_models(
-            repo_id=smoldocling_vlm_conversion_options.repo_id,
-            local_dir=output_dir / smoldocling_vlm_conversion_options.repo_cache_folder,
+        download_hf_model(
+            repo_id=SMOLDOCLING_TRANSFORMERS.repo_id,
+            local_dir=output_dir / SMOLDOCLING_TRANSFORMERS.repo_cache_folder,
             force=force,
             progress=progress,
         )
 
     if with_smoldocling_mlx:
         _log.info("Downloading SmolDocling MLX model...")
-        HuggingFaceVlmModel.download_models(
-            repo_id=smoldocling_vlm_mlx_conversion_options.repo_id,
-            local_dir=output_dir
-            / smoldocling_vlm_mlx_conversion_options.repo_cache_folder,
+        download_hf_model(
+            repo_id=SMOLDOCLING_MLX.repo_id,
+            local_dir=output_dir / SMOLDOCLING_MLX.repo_cache_folder,
             force=force,
             progress=progress,
         )
 
     if with_granite_vision:
         _log.info("Downloading Granite Vision model...")
-        PictureDescriptionVlmModel.download_models(
+        download_hf_model(
             repo_id=granite_picture_description.repo_id,
             local_dir=output_dir / granite_picture_description.repo_cache_folder,
             force=force,

{docling-2.34.0.dist-info → docling-2.36.0.dist-info}/METADATA

@@ -1,67 +1,68 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.4
 Name: docling
-Version: 2.34.0
+Version: 2.36.0
 Summary: SDK and CLI for parsing PDF, DOCX, HTML, and more, to a unified document representation for powering downstream workflows such as gen AI applications.
-Home-page: https://github.com/docling-project/docling
-License: MIT
+Author-email: Christoph Auer <cau@zurich.ibm.com>, Michele Dolfi <dol@zurich.ibm.com>, Maxim Lysak <mly@zurich.ibm.com>, Nikos Livathinos <nli@zurich.ibm.com>, Ahmed Nassar <ahn@zurich.ibm.com>, Panos Vagenas <pva@zurich.ibm.com>, Peter Staar <taa@zurich.ibm.com>
+License-Expression: MIT
+Project-URL: homepage, https://github.com/docling-project/docling
+Project-URL: repository, https://github.com/docling-project/docling
+Project-URL: issues, https://github.com/docling-project/docling/issues
+Project-URL: changelog, https://github.com/docling-project/docling/blob/main/CHANGELOG.md
 Keywords: docling,convert,document,pdf,docx,html,markdown,layout model,segmentation,table structure,table former
-Author: Christoph Auer
-Author-email: cau@zurich.ibm.com
-Requires-Python: >=3.9,<4.0
+Classifier: Operating System :: MacOS :: MacOS X
+Classifier: Operating System :: POSIX :: Linux
+Classifier: Operating System :: Microsoft :: Windows
 Classifier: Development Status :: 5 - Production/Stable
 Classifier: Intended Audience :: Developers
 Classifier: Intended Audience :: Science/Research
-Classifier: License :: OSI Approved :: MIT License
-Classifier: Operating System :: MacOS :: MacOS X
-Classifier: Operating System :: POSIX :: Linux
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
 Classifier: Programming Language :: Python :: 3
 Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
-Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
-Provides-Extra: ocrmac
-Provides-Extra: rapidocr
+Requires-Python: <4.0,>=3.9
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: pydantic<3.0.0,>=2.0.0
+Requires-Dist: docling-core[chunking]<3.0.0,>=2.29.0
+Requires-Dist: docling-ibm-models<4.0.0,>=3.4.4
+Requires-Dist: docling-parse<5.0.0,>=4.0.0
+Requires-Dist: filetype<2.0.0,>=1.2.0
+Requires-Dist: pypdfium2<5.0.0,>=4.30.0
+Requires-Dist: pydantic-settings<3.0.0,>=2.3.0
+Requires-Dist: huggingface_hub<1,>=0.23
+Requires-Dist: requests<3.0.0,>=2.32.2
+Requires-Dist: easyocr<2.0,>=1.7
+Requires-Dist: certifi>=2024.7.4
+Requires-Dist: rtree<2.0.0,>=1.3.0
+Requires-Dist: typer<0.16.0,>=0.12.5
+Requires-Dist: python-docx<2.0.0,>=1.1.2
+Requires-Dist: python-pptx<2.0.0,>=1.0.2
+Requires-Dist: beautifulsoup4<5.0.0,>=4.12.3
+Requires-Dist: pandas<3.0.0,>=2.1.4
+Requires-Dist: marko<3.0.0,>=2.1.2
+Requires-Dist: openpyxl<4.0.0,>=3.1.5
+Requires-Dist: lxml<6.0.0,>=4.0.0
+Requires-Dist: pillow<12.0.0,>=10.0.0
+Requires-Dist: tqdm<5.0.0,>=4.65.0
+Requires-Dist: pluggy<2.0.0,>=1.0.0
+Requires-Dist: pylatexenc<3.0,>=2.10
+Requires-Dist: click<8.2.0
+Requires-Dist: scipy<2.0.0,>=1.6.0
 Provides-Extra: tesserocr
+Requires-Dist: tesserocr<3.0.0,>=2.7.1; extra == "tesserocr"
+Provides-Extra: ocrmac
+Requires-Dist: ocrmac<2.0.0,>=1.0.0; sys_platform == "darwin" and extra == "ocrmac"
 Provides-Extra: vlm
-Requires-Dist: accelerate (>=1.2.1,<2.0.0) ; (sys_platform != "darwin" or platform_machine != "x86_64") and (extra == "vlm")
-Requires-Dist: beautifulsoup4 (>=4.12.3,<5.0.0)
-Requires-Dist: certifi (>=2024.7.4)
-Requires-Dist: click (<8.2.0)
-Requires-Dist: docling-core[chunking] (>=2.29.0,<3.0.0)
-Requires-Dist: docling-ibm-models (>=3.4.0,<4.0.0)
-Requires-Dist: docling-parse (>=4.0.0,<5.0.0)
-Requires-Dist: easyocr (>=1.7,<2.0)
-Requires-Dist: filetype (>=1.2.0,<2.0.0)
-Requires-Dist: huggingface_hub (>=0.23,<1)
-Requires-Dist: lxml (>=4.0.0,<6.0.0)
-Requires-Dist: marko (>=2.1.2,<3.0.0)
-Requires-Dist: ocrmac (>=1.0.0,<2.0.0) ; (sys_platform == "darwin") and (extra == "ocrmac")
-Requires-Dist: onnxruntime (>=1.7.0,<1.20.0) ; (python_version < "3.10") and (extra == "rapidocr")
-Requires-Dist: onnxruntime (>=1.7.0,<2.0.0) ; (python_version >= "3.10") and (extra == "rapidocr")
-Requires-Dist: openpyxl (>=3.1.5,<4.0.0)
-Requires-Dist: pandas (>=2.1.4,<3.0.0)
-Requires-Dist: pillow (>=10.0.0,<12.0.0)
-Requires-Dist: pluggy (>=1.0.0,<2.0.0)
-Requires-Dist: pydantic (>=2.0.0,<3.0.0)
-Requires-Dist: pydantic-settings (>=2.3.0,<3.0.0)
-Requires-Dist: pylatexenc (>=2.10,<3.0)
-Requires-Dist: pypdfium2 (>=4.30.0,<5.0.0)
-Requires-Dist: python-docx (>=1.1.2,<2.0.0)
-Requires-Dist: python-pptx (>=1.0.2,<2.0.0)
-Requires-Dist: rapidocr-onnxruntime (>=1.4.0,<2.0.0) ; (python_version < "3.13") and (extra == "rapidocr")
-Requires-Dist: requests (>=2.32.2,<3.0.0)
-Requires-Dist: rtree (>=1.3.0,<2.0.0)
-Requires-Dist: scipy (>=1.6.0,<1.14.0) ; python_version < "3.10"
-Requires-Dist: scipy (>=1.6.0,<2.0.0) ; python_version >= "3.10"
-Requires-Dist: tesserocr (>=2.7.1,<3.0.0) ; extra == "tesserocr"
-Requires-Dist: tqdm (>=4.65.0,<5.0.0)
-Requires-Dist: transformers (>=4.42.0,<4.43.0) ; (sys_platform == "darwin" and platform_machine == "x86_64") and (extra == "vlm")
-Requires-Dist: transformers (>=4.46.0,<5.0.0) ; (sys_platform != "darwin" or platform_machine != "x86_64") and (extra == "vlm")
-Requires-Dist: typer (>=0.12.5,<0.16.0)
-Project-URL: Repository, https://github.com/docling-project/docling
-Description-Content-Type: text/markdown
+Requires-Dist: transformers<5.0.0,>=4.46.0; extra == "vlm"
+Requires-Dist: accelerate<2.0.0,>=1.2.1; extra == "vlm"
+Requires-Dist: mlx-vlm>=0.1.22; (python_version >= "3.10" and sys_platform == "darwin" and platform_machine == "arm64") and extra == "vlm"
+Provides-Extra: rapidocr
+Requires-Dist: rapidocr-onnxruntime<2.0.0,>=1.4.0; python_version < "3.13" and extra == "rapidocr"
+Requires-Dist: onnxruntime<2.0.0,>=1.7.0; extra == "rapidocr"
+Dynamic: license-file
 
 <p align="center">
   <a href="https://github.com/docling-project/docling">
@@ -79,9 +80,8 @@ Description-Content-Type: text/markdown
 [![Docs](https://img.shields.io/badge/docs-live-brightgreen)](https://docling-project.github.io/docling/)
 [![PyPI version](https://img.shields.io/pypi/v/docling)](https://pypi.org/project/docling/)
 [![PyPI - Python Version](https://img.shields.io/pypi/pyversions/docling)](https://pypi.org/project/docling/)
-[![Poetry](https://img.shields.io/endpoint?url=https://python-poetry.org/badge/v0.json)](https://python-poetry.org/)
-[![Code style: black](https://img.shields.io/badge/code%20style-black-000000.svg)](https://github.com/psf/black)
-[![Imports: isort](https://img.shields.io/badge/%20imports-isort-%231674b1?style=flat&labelColor=ef8336)](https://pycqa.github.io/isort/)
+[![uv](https://img.shields.io/endpoint?url=https://raw.githubusercontent.com/astral-sh/uv/main/assets/badge/v0.json)](https://github.com/astral-sh/uv)
+[![Ruff](https://img.shields.io/endpoint?url=https://raw.githubusercontent.com/astral-sh/ruff/main/assets/badge/v2.json)](https://github.com/astral-sh/ruff)
 [![Pydantic v2](https://img.shields.io/endpoint?url=https://raw.githubusercontent.com/pydantic/pydantic/main/docs/badge/v2.json)](https://pydantic.dev)
 [![pre-commit](https://img.shields.io/badge/pre--commit-enabled-brightgreen?logo=pre-commit&logoColor=white)](https://github.com/pre-commit/pre-commit)
 [![License MIT](https://img.shields.io/github/license/docling-project/docling)](https://opensource.org/licenses/MIT)
@@ -101,7 +101,7 @@ Docling simplifies document processing, parsing diverse formats — including ad
 * 🔒 Local execution capabilities for sensitive data and air-gapped environments
 * 🤖 Plug-and-play [integrations][integrations] incl. LangChain, LlamaIndex, Crew AI & Haystack for agentic AI
 * 🔍 Extensive OCR support for scanned PDFs and images
-* 🥚 Support of Visual Language Models ([SmolDocling](https://huggingface.co/ds4sd/SmolDocling-256M-preview)) 🆕
+* 🥚 Support of several Visual Language Models ([SmolDocling](https://huggingface.co/ds4sd/SmolDocling-256M-preview))
 * 💻 Simple and convenient CLI
 
 ### Coming soon
@@ -214,4 +214,3 @@ The project was started by the AI for knowledge team at IBM Research Zurich.
 [supported_formats]: https://docling-project.github.io/docling/usage/supported_formats/
 [docling_document]: https://docling-project.github.io/docling/concepts/docling_document/
 [integrations]: https://docling-project.github.io/docling/integrations/
-