Adding a minimal haystack gpu build (#2185)

* added minimal gpu image

* Update Documentation & Code Style

* removed old installations

* build minimal images in gh action

* quotes to single quotes

* switched repos

* fix

* fix ordering

* move to deepset dockerhub acc

* update nvidia/cuda image to match newest torch+cu

* refactor Dockerfile-GPU-minimal to optimize build

* Remove spurious doc changes from tutorial6

* install ocr dependencies and pdftotext

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: Sara Zan <sarazanzo94@gmail.com>
Co-authored-by: Ivan Lopez <ivan@askai.net>

.github/workflows/docker_build.yml

@@ -16,9 +16,18 @@ jobs:
         include:
           - dockerfile: Dockerfile
             repository: deepset/haystack-cpu
+            tagprefix: ''
+            default: true
 
           - dockerfile: Dockerfile-GPU
             repository: deepset/haystack-gpu
+            tagprefix: 'demo-'
+            default: true
+
+          - dockerfile: Dockerfile-GPU-minimal
+            repository: deepset/haystack-gpu
+            tagprefix: 'minimal-'
+            default: false
 
     steps:
       - name: Checkout
@@ -30,11 +39,12 @@ jobs:
         with:
           images: |
             ${{ matrix.repository }}
-          flavor: |
-            latest=true
           tags: |
-            type=ref,event=tag
-            type=sha,format=long,prefix=
+            type=ref,event=tag,prefix=${{ matrix.tagprefix }}                     
+            type=sha,format=long,prefix=${{ matrix.tagprefix }}
+            type=sha,format=long,prefix=,enable=${{ matrix.default }}
+            type=raw,value=latest,prefix=${{ matrix.tagprefix }}
+            type=raw,value=latest,prefix=,enable=${{ matrix.default }}
 
       - name: Set up Docker Buildx
         uses: docker/setup-buildx-action@v1

Dockerfile-GPU-minimal

@@ -0,0 +1,40 @@
+FROM nvidia/cuda:11.3.1-runtime-ubuntu20.04
+
+WORKDIR /home/user
+
+ENV LC_ALL=C.UTF-8
+ENV LANG=C.UTF-8
+
+# Install software dependencies
+RUN apt-get update && apt-get install -y software-properties-common && \
+    add-apt-repository ppa:deadsnakes/ppa && \
+    DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \
+        curl \
+        git \
+        poppler-utils \
+        python3-pip \
+        python3.8 \
+        python3.8-distutils \
+        swig \
+        tesseract-ocr && \
+    # Cleanup apt cache
+    rm -rf /var/lib/apt/lists/* && \
+    # Install PDF converter
+    curl -s https://dl.xpdfreader.com/xpdf-tools-linux-4.03.tar.gz \
+    | tar -xvzf - -C /usr/local/bin --strip-components=2 xpdf-tools-linux-4.03/bin64/pdftotext
+
+# Copy Haystack package files but not the source code
+COPY setup.py setup.cfg pyproject.toml VERSION.txt LICENSE README.md /home/user/
+
+# Install all the dependencies, including ocr component
+RUN pip3 install --upgrade --no-cache-dir pip && \
+    pip3 install --no-cache-dir .[ocr] && \
+    # Install PyTorch with CUDA 11
+    pip3 install --no-cache-dir torch==1.10.2+cu113 -f https://download.pytorch.org/whl/torch_stable.html
+
+# Copy Haystack source code
+COPY haystack /home/user/haystack/
+# Install Haystack
+RUN pip3 install --no-cache-dir --no-deps .[ocr] && \
+    # Cleanup copied files after installation is completed
+    rm -rf /home/user/*