feat: package updates and more cuda images (#229)

Signed-off-by: Michele Dolfi <dol@zurich.ibm.com>
2025-11-29 00:23:36 +00:00 · 2025-06-24 09:59:05 -05:00
parent 717fb3a8d8
commit 30aca92298
13 changed files with 2151 additions and 1489 deletions
--- a/.github/workflows/ci-images-dryrun.yml
+++ b/.github/workflows/ci-images-dryrun.yml
@@ -15,15 +15,23 @@ jobs:
        spec:
          - name: docling-project/docling-serve
            build_args: |
-              UV_SYNC_EXTRA_ARGS=--no-extra cu124 --no-extra cpu
+              UV_SYNC_EXTRA_ARGS=--no-extra flash-attn
            platforms: linux/amd64, linux/arm64
          - name: docling-project/docling-serve-cpu
            build_args: |
-              UV_SYNC_EXTRA_ARGS=--no-extra cu124 --no-extra flash-attn
+              UV_SYNC_EXTRA_ARGS=--no-group pypi --group cpu --no-extra flash-attn
            platforms: linux/amd64, linux/arm64
          - name: docling-project/docling-serve-cu124
            build_args: |
-              UV_SYNC_EXTRA_ARGS=--no-extra cpu
+              UV_SYNC_EXTRA_ARGS=--no-group pypi --group cu124
+            platforms: linux/amd64
+          - name: docling-project/docling-serve-cu126
+            build_args: |
+              UV_SYNC_EXTRA_ARGS=--no-group pypi --group cu126
+            platforms: linux/amd64
+          - name: docling-project/docling-serve-cu128
+            build_args: |
+              UV_SYNC_EXTRA_ARGS=--no-group pypi --group cu128
            platforms: linux/amd64

    permissions:
--- a/.github/workflows/images.yml
+++ b/.github/workflows/images.yml
@@ -19,15 +19,23 @@ jobs:
        spec:
          - name: docling-project/docling-serve
            build_args: |
-              UV_SYNC_EXTRA_ARGS=--no-extra cu124 --no-extra cpu
+              UV_SYNC_EXTRA_ARGS=--no-extra flash-attn
            platforms: linux/amd64, linux/arm64
          - name: docling-project/docling-serve-cpu
            build_args: |
-              UV_SYNC_EXTRA_ARGS=--no-extra cu124 --no-extra flash-attn
+              UV_SYNC_EXTRA_ARGS=--no-group pypi --group cpu --no-extra flash-attn
            platforms: linux/amd64, linux/arm64
          - name: docling-project/docling-serve-cu124
            build_args: |
-              UV_SYNC_EXTRA_ARGS=--no-extra cpu
+              UV_SYNC_EXTRA_ARGS=--no-group pypi --group cu124
+            platforms: linux/amd64
+          - name: docling-project/docling-serve-cu126
+            build_args: |
+              UV_SYNC_EXTRA_ARGS=--no-group pypi --group cu126
+            platforms: linux/amd64
+          - name: docling-project/docling-serve-cu128
+            build_args: |
+              UV_SYNC_EXTRA_ARGS=--no-group pypi --group cu128
            platforms: linux/amd64

    permissions:
--- a/.github/workflows/job-build.yml
+++ b/.github/workflows/job-build.yml
@@ -17,7 +17,7 @@ jobs:
          python-version: ${{ matrix.python-version }}
          enable-cache: true
      - name: Install dependencies
-        run: uv sync --all-extras --no-extra cu124 --no-extra flash-attn
+        run: uv sync --all-extras --no-extra flash-attn
      - name: Build package
        run: uv build
      - name: Check content of wheel
--- a/.github/workflows/job-checks.yml
+++ b/.github/workflows/job-checks.yml
@@ -25,7 +25,7 @@ jobs:
          key: pre-commit|${{ env.PY }}|${{ hashFiles('.pre-commit-config.yaml') }}

      - name: Install dependencies
-        run: uv sync --frozen --all-extras --no-extra cu124 --no-extra flash-attn
+        run: uv sync --frozen --all-extras --no-extra flash-attn

      - name: Run styling check
        run: pre-commit run --all-files
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@@ -23,6 +23,6 @@ repos:
        files: '\.py$'
  - repo: https://github.com/astral-sh/uv-pre-commit
    # uv version.
-    rev: 0.6.1
+    rev: 0.7.13
    hooks:
      - id: uv-lock
--- a/4
+++ b/4
@@ -42,7 +42,7 @@ ENV \

 ARG UV_SYNC_EXTRA_ARGS=""

-RUN --mount=from=ghcr.io/astral-sh/uv:0.6.1,source=/uv,target=/bin/uv \
+RUN --mount=from=ghcr.io/astral-sh/uv:0.7.13,source=/uv,target=/bin/uv \
    --mount=type=cache,target=/opt/app-root/src/.cache/uv,uid=1001 \
    --mount=type=bind,source=uv.lock,target=uv.lock \
    --mount=type=bind,source=pyproject.toml,target=pyproject.toml \
@@ -61,7 +61,7 @@ RUN echo "Downloading models..." && \
    chmod -R g=u ${DOCLING_SERVE_ARTIFACTS_PATH}

 COPY --chown=1001:0 ./docling_serve ./docling_serve
-RUN --mount=from=ghcr.io/astral-sh/uv:0.6.1,source=/uv,target=/bin/uv \
+RUN --mount=from=ghcr.io/astral-sh/uv:0.7.13,source=/uv,target=/bin/uv \
    --mount=type=cache,target=/opt/app-root/src/.cache/uv,uid=1001 \
    --mount=type=bind,source=uv.lock,target=uv.lock \
    --mount=type=bind,source=pyproject.toml,target=pyproject.toml \
--- a/34
+++ b/34
@@ -26,26 +26,40 @@ md-lint-file:
 	$(CMD_PREFIX) touch .markdown-lint

 .PHONY: docling-serve-image
-docling-serve-image: Containerfile
+docling-serve-image: Containerfile ## Build docling-serve container image
 	$(ECHO_PREFIX) printf "  %-12s Containerfile\n" "[docling-serve]"
-	$(CMD_PREFIX) docker build --load --build-arg "UV_SYNC_EXTRA_ARGS=--no-extra cu124 --no-extra cpu" -f Containerfile -t ghcr.io/docling-project/docling-serve:$(TAG) .
+	$(CMD_PREFIX) docker build --load -f Containerfile -t ghcr.io/docling-project/docling-serve:$(TAG) .
 	$(CMD_PREFIX) docker tag ghcr.io/docling-project/docling-serve:$(TAG) ghcr.io/docling-project/docling-serve:$(BRANCH_TAG)
 	$(CMD_PREFIX) docker tag ghcr.io/docling-project/docling-serve:$(TAG) quay.io/docling-project/docling-serve:$(BRANCH_TAG)

 .PHONY: docling-serve-cpu-image
 docling-serve-cpu-image: Containerfile ## Build docling-serve "cpu only" container image
 	$(ECHO_PREFIX) printf "  %-12s Containerfile\n" "[docling-serve CPU]"
-	$(CMD_PREFIX) docker build --load --build-arg "UV_SYNC_EXTRA_ARGS=--no-extra cu124 --no-extra flash-attn" -f Containerfile -t ghcr.io/docling-project/docling-serve-cpu:$(TAG) .
+	$(CMD_PREFIX) docker build --load --build-arg "UV_SYNC_EXTRA_ARGS=--no-group pypi --group cpu --no-extra flash-attn" -f Containerfile -t ghcr.io/docling-project/docling-serve-cpu:$(TAG) .
 	$(CMD_PREFIX) docker tag ghcr.io/docling-project/docling-serve-cpu:$(TAG) ghcr.io/docling-project/docling-serve-cpu:$(BRANCH_TAG)
 	$(CMD_PREFIX) docker tag ghcr.io/docling-project/docling-serve-cpu:$(TAG) quay.io/docling-project/docling-serve-cpu:$(BRANCH_TAG)

 .PHONY: docling-serve-cu124-image
-docling-serve-cu124-image: Containerfile ## Build docling-serve container image with GPU support
+docling-serve-cu124-image: Containerfile ## Build docling-serve container image with CUDA 12.4 support
 	$(ECHO_PREFIX) printf "  %-12s Containerfile\n" "[docling-serve with Cuda 12.4]"
-	$(CMD_PREFIX) docker build --load --build-arg "UV_SYNC_EXTRA_ARGS=--no-extra cpu" -f Containerfile --platform linux/amd64 -t ghcr.io/docling-project/docling-serve-cu124:$(TAG) .
+	$(CMD_PREFIX) docker build --load --build-arg "UV_SYNC_EXTRA_ARGS=--no-group pypi --group cu124" -f Containerfile --platform linux/amd64 -t ghcr.io/docling-project/docling-serve-cu124:$(TAG) .
 	$(CMD_PREFIX) docker tag ghcr.io/docling-project/docling-serve-cu124:$(TAG) ghcr.io/docling-project/docling-serve-cu124:$(BRANCH_TAG)
 	$(CMD_PREFIX) docker tag ghcr.io/docling-project/docling-serve-cu124:$(TAG) quay.io/docling-project/docling-serve-cu124:$(BRANCH_TAG)

+.PHONY: docling-serve-cu126-image
+docling-serve-cu126-image: Containerfile ## Build docling-serve container image with CUDA 12.6 support
+	$(ECHO_PREFIX) printf "  %-12s Containerfile\n" "[docling-serve with Cuda 12.6]"
+	$(CMD_PREFIX) docker build --load --build-arg "UV_SYNC_EXTRA_ARGS=--no-group pypi --group cu126" -f Containerfile --platform linux/amd64 -t ghcr.io/docling-project/docling-serve-cu126:$(TAG) .
+	$(CMD_PREFIX) docker tag ghcr.io/docling-project/docling-serve-cu126:$(TAG) ghcr.io/docling-project/docling-serve-cu126:$(BRANCH_TAG)
+	$(CMD_PREFIX) docker tag ghcr.io/docling-project/docling-serve-cu126:$(TAG) quay.io/docling-project/docling-serve-cu126:$(BRANCH_TAG)
+
+.PHONY: docling-serve-cu128-image
+docling-serve-cu128-image: Containerfile ## Build docling-serve container image with CUDA 12.8 support
+	$(ECHO_PREFIX) printf "  %-12s Containerfile\n" "[docling-serve with Cuda 12.8]"
+	$(CMD_PREFIX) docker build --load --build-arg "UV_SYNC_EXTRA_ARGS=--no-group pypi --group cu128" -f Containerfile --platform linux/amd64 -t ghcr.io/docling-project/docling-serve-cu128:$(TAG) .
+	$(CMD_PREFIX) docker tag ghcr.io/docling-project/docling-serve-cu128:$(TAG) ghcr.io/docling-project/docling-serve-cu128:$(BRANCH_TAG)
+	$(CMD_PREFIX) docker tag ghcr.io/docling-project/docling-serve-cu128:$(TAG) quay.io/docling-project/docling-serve-cu128:$(BRANCH_TAG)
+
 .PHONY: action-lint
 action-lint: .action-lint ##      Lint GitHub Action workflows
 .action-lint: $(shell find .github -type f) | action-lint-file
@@ -87,9 +101,9 @@ run-docling-cpu: ## Run the docling-serve container with CPU support and assign
 	$(ECHO_PREFIX) printf "  %-12s Running docling-serve container with CPU support on port 5001...\n" "[RUN CPU]"
 	$(CMD_PREFIX) docker run -it --name docling-serve-cpu -p 5001:5001 ghcr.io/docling-project/docling-serve-cpu:main

-.PHONY: run-docling-gpu
-run-docling-gpu: ## Run the docling-serve container with GPU support and assign a container name
+.PHONY: run-docling-cu124
+run-docling-cu124: ## Run the docling-serve container with GPU support and assign a container name
 	$(ECHO_PREFIX) printf "  %-12s Removing existing container if it exists...\n" "[CLEANUP]"
-	$(CMD_PREFIX) docker rm -f docling-serve-gpu 2>/dev/null || true
-	$(ECHO_PREFIX) printf "  %-12s Running docling-serve container with GPU support on port 5001...\n" "[RUN GPU]"
-	$(CMD_PREFIX) docker run -it --name docling-serve-gpu -p 5001:5001 ghcr.io/docling-project/docling-serve:main
+	$(CMD_PREFIX) docker rm -f docling-serve-cu124 2>/dev/null || true
+	$(ECHO_PREFIX) printf "  %-12s Running docling-serve container with GPU support on port 5001...\n" "[RUN CUDA 12.4]"
+	$(CMD_PREFIX) docker run -it --name docling-serve-cu124 -p 5001:5001 ghcr.io/docling-project/docling-serve-cu124:main
--- a/README.md
+++ b/README.md
@@ -45,9 +45,11 @@ Available container images:

 | Name | Description | Arch | Size |
 | -----|-------------|------|------|
-| [`ghcr.io/docling-project/docling-serve`](https://github.com/docling-project/docling-serve/pkgs/container/docling-serve) <br /> [`quay.io/docling-project/docling-serve`](https://quay.io/repository/docling-project/docling-serve) | Simple image for Docling Serve, installing all packages from the official pypi.org index. | `linux/amd64`, `linux/arm64` | 3.6 GB |
+| [`ghcr.io/docling-project/docling-serve`](https://github.com/docling-project/docling-serve/pkgs/container/docling-serve) <br /> [`quay.io/docling-project/docling-serve`](https://quay.io/repository/docling-project/docling-serve) | Simple image for Docling Serve, installing all packages from the official pypi.org index. | `linux/amd64`, `linux/arm64` | 3.6 GB (arm64) <br /> 8.7 GB (amd64) |
 | [`ghcr.io/docling-project/docling-serve-cpu`](https://github.com/docling-project/docling-serve/pkgs/container/docling-serve-cpu) <br /> [`quay.io/docling-project/docling-serve-cpu`](https://quay.io/repository/docling-project/docling-serve-cpu) | Cpu-only image which installs `torch` from the pytorch cpu index. | `linux/amd64`, `linux/arm64` | 3.6 GB |
 | [`ghcr.io/docling-project/docling-serve-cu124`](https://github.com/docling-project/docling-serve/pkgs/container/docling-serve-cu124) <br /> [`quay.io/docling-project/docling-serve-cu124`](https://quay.io/repository/docling-project/docling-serve-cu124) | Cuda 12.4 image which installs `torch` from the pytorch cu124 index. | `linux/amd64` | 8.7 GB |
+| [`ghcr.io/docling-project/docling-serve-cu126`](https://github.com/docling-project/docling-serve/pkgs/container/docling-serve-cu126) <br /> [`quay.io/docling-project/docling-serve-cu126`](https://quay.io/repository/docling-project/docling-serve-cu126) | Cuda 12.6 image which installs `torch` from the pytorch cu126 index. | `linux/amd64` | 8.7 GB |
+| [`ghcr.io/docling-project/docling-serve-cu128`](https://github.com/docling-project/docling-serve/pkgs/container/docling-serve-cu128) <br /> [`quay.io/docling-project/docling-serve-cu128`](https://quay.io/repository/docling-project/docling-serve-cu128) | Cuda 12.8 image which installs `torch` from the pytorch cu128 index. | `linux/amd64` | 8.7 GB |

 Coming soon: `docling-serve-slim` images will reduce the size by skipping the model weights download.

--- a/docling_serve/datamodel/convert.py
+++ b/docling_serve/datamodel/convert.py
@@ -8,8 +8,8 @@ from docling.datamodel.base_models import InputFormat, OutputFormat
 from docling.datamodel.pipeline_options import (
    EasyOcrOptions,
    PdfBackend,
-    PdfPipeline,
    PictureDescriptionBaseOptions,
+    ProcessingPipeline,
    TableFormerMode,
    TableStructureOptions,
 )
@@ -227,9 +227,9 @@ class ConvertDocumentsOptions(BaseModel):
    ] = TableStructureOptions().mode

    pipeline: Annotated[
-        PdfPipeline,
+        ProcessingPipeline,
        Field(description="Choose the pipeline to process PDF or image files."),
-    ] = PdfPipeline.STANDARD
+    ] = ProcessingPipeline.STANDARD

    page_range: Annotated[
        PageRange,
--- a/docling_serve/docling_conversion.py
+++ b/docling_serve/docling_conversion.py
@@ -19,10 +19,10 @@ from docling.datamodel.document import ConversionResult
 from docling.datamodel.pipeline_options import (
    OcrOptions,
    PdfBackend,
-    PdfPipeline,
    PdfPipelineOptions,
    PictureDescriptionApiOptions,
    PictureDescriptionVlmOptions,
+    ProcessingPipeline,
    TableFormerMode,
    VlmPipelineOptions,
    smoldocling_vlm_conversion_options,
@@ -217,7 +217,7 @@ def get_pdf_pipeline_opts(
        )

    pipeline_options: Union[PdfPipelineOptions, VlmPipelineOptions]
-    if request.pipeline == PdfPipeline.STANDARD:
+    if request.pipeline == ProcessingPipeline.STANDARD:
        pipeline_options = _parse_standard_pdf_opts(request, artifacts_path)
        backend = _parse_backend(request)
        pdf_format_option = PdfFormatOption(
@@ -225,7 +225,7 @@ def get_pdf_pipeline_opts(
            backend=backend,
        )

-    elif request.pipeline == PdfPipeline.VLM:
+    elif request.pipeline == ProcessingPipeline.VLM:
        pipeline_options = _parse_vlm_pdf_opts(request, artifacts_path)
        pdf_format_option = PdfFormatOption(
            pipeline_cls=VlmPipeline, pipeline_options=pipeline_options
--- a/docling_serve/gradio_ui.py
+++ b/docling_serve/gradio_ui.py
@@ -16,7 +16,7 @@ import httpx
 from docling.datamodel.base_models import FormatToExtensions
 from docling.datamodel.pipeline_options import (
    PdfBackend,
-    PdfPipeline,
+    ProcessingPipeline,
    TableFormerMode,
    TableStructureOptions,
 )
@@ -587,9 +587,9 @@ with gr.Blocks(
        with gr.Row():
            with gr.Column(scale=1, min_width=200):
                pipeline = gr.Radio(
-                    [(v.value.capitalize(), v.value) for v in PdfPipeline],
+                    [(v.value.capitalize(), v.value) for v in ProcessingPipeline],
                    label="Pipeline type",
-                    value=PdfPipeline.STANDARD.value,
+                    value=ProcessingPipeline.STANDARD.value,
                )
        with gr.Row():
            with gr.Column(scale=1, min_width=200):
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -30,7 +30,7 @@ classifiers = [
 ]
 requires-python = ">=3.10"
 dependencies = [
-    "docling[vlm]~=2.28",
+    "docling[vlm]~=2.38",
    "docling-core>=2.32.0",
    "mlx-vlm~=0.1.12; sys_platform == 'darwin' and platform_machine == 'arm64'",
    "fastapi[standard]~=0.115",
@@ -57,14 +57,6 @@ rapidocr = [
    "rapidocr-onnxruntime~=1.4; python_version<'3.13'",
    "onnxruntime~=1.7",
 ]
-cpu = [
-  "torch>=2.6.0",
-  "torchvision>=0.21.0",
-]
-cu124 = [
-  "torch>=2.6.0",
-  "torchvision>=0.21.0",
-]
 flash-attn = [
  "flash-attn~=2.7.0; sys_platform == 'linux' and platform_machine == 'x86_64'"
 ]
@@ -80,18 +72,39 @@ dev = [
    "python-semantic-release~=7.32",
    "ruff>=0.9.6",
 ]
+pypi = [
+  "torch>=2.6.0",
+  "torchvision>=0.21.0",
+]
+cpu = [
+  "torch>=2.6.0",
+  "torchvision>=0.21.0",
+]
+cu124 = [
+  "torch>=2.6.0",
+  "torchvision>=0.21.0",
+]
+cu126 = [
+  "torch>=2.6.0",
+  "torchvision>=0.21.0",
+]
+cu128 = [
+  "torch>=2.7.0",
+  "torchvision>=0.22.0",
+]

 [tool.uv]
 package = true
+default-groups = ["dev", "pypi"]
 conflicts = [
  [
-    { extra = "cpu" },
-    { extra = "cu124" },
+    { group = "pypi" },
+    { group = "cpu" },
+    { group = "cu124" },
+    { group = "cu126" },
+    { group = "cu128" },
  ],
-  [
-    { extra = "cpu" },
-    { extra = "flash-attn" },
-  ],]
+]
 environments = ["sys_platform != 'darwin' or platform_machine != 'x86_64'"]
 override-dependencies = [
  "urllib3~=2.0"
@@ -99,14 +112,25 @@ override-dependencies = [

 [tool.uv.sources]
 torch = [
-  { index = "pytorch-cpu", extra = "cpu" },
-  { index = "pytorch-cu124", extra = "cu124" },
+  { index = "pytorch-pypi", group = "pypi" },
+  { index = "pytorch-cpu", group = "cpu" },
+  { index = "pytorch-cu124", group = "cu124" },
+  { index = "pytorch-cu126", group = "cu126" },
+  { index = "pytorch-cu128", group = "cu128" },
 ]
 torchvision = [
-  { index = "pytorch-cpu", extra = "cpu" },
-  { index = "pytorch-cu124", extra = "cu124" },
+  { index = "pytorch-pypi", group = "pypi" },
+  { index = "pytorch-cpu", group = "cpu" },
+  { index = "pytorch-cu124", group = "cu124" },
+  { index = "pytorch-cu126", group = "cu126" },
+  { index = "pytorch-cu128", group = "cu128" },
 ]

+[[tool.uv.index]]
+name = "pytorch-pypi"
+url = "https://pypi.org/simple"
+explicit = true
+
 [[tool.uv.index]]
 name = "pytorch-cpu"
 url = "https://download.pytorch.org/whl/cpu"
@@ -117,6 +141,16 @@ name = "pytorch-cu124"
 url = "https://download.pytorch.org/whl/cu124"
 explicit = true

+[[tool.uv.index]]
+name = "pytorch-cu126"
+url = "https://download.pytorch.org/whl/cu126"
+explicit = true
+
+[[tool.uv.index]]
+name = "pytorch-cu128"
+url = "https://download.pytorch.org/whl/cu128"
+explicit = true
+
 [tool.setuptools.packages.find]
 include = ["docling_serve*"]
 namespaces = true
--- a/uv.lock
+++ b/uv.lock