Merge pull request #190 from zenml-io/sandbox/add-dockerfile-for-omni-reader

Add Dockerfile for running OmniReader in sandbox environment

Author: marwan37

omni-reader/Dockerfile.sandbox

@@ -0,0 +1,44 @@
+# Sandbox base image
+FROM safoinext/zenml-sandbox:latest
+
+# Install project-specific dependencies
+# Install polars-lts-cpu instead of polars (version compiled for CPU compatibility)
+RUN pip install --no-cache-dir \
+    "instructor==1.7.7" \
+    "jiwer==3.0.5" \
+    "jiter==0.8.2" \
+    "importlib-metadata<7.0,>=1.4.0" \
+    "litellm==1.64.1" \
+    "mistralai==1.0.3" \
+    "numpy<2.0,>=1.9.0" \
+    "openai==1.69.0" \
+    "Pillow==11.1.0" \
+    "polars-lts-cpu==1.26.0" \
+    "pyarrow>=7.0.0" \
+    "python-dotenv==1.0.1" \
+    "streamlit==1.44.0" \
+    "pydantic>=2.8.2,<2.9.0" \
+    "tqdm==4.66.4" \
+    "zenml>=0.80.0" \
+    uv
+
+# Set workspace directory
+WORKDIR /workspace
+
+# Clone only the omni-reader directory and reorganize
+RUN git clone --depth 1 https://github.com/zenml-io/zenml-projects.git /tmp/zenml-projects && \
+    cp -r /tmp/zenml-projects/omni-reader/* /workspace/ && \
+    rm -rf /tmp/zenml-projects
+
+# Create a template .env file for API keys
+RUN echo "OPENAI_API_KEY=YOUR_OPENAI_API_KEY_HERE" > .env && \
+    echo "MISTRAL_API_KEY=YOUR_MISTRAL_API_KEY_HERE" >> .env
+
+# Create a settings directory (mainly to auto-apply the dark theme)
+RUN mkdir -p /root/.local/share/code-server/User
+# Copy settings file
+COPY settings.json /root/.local/share/code-server/User/
+
+# Set environment variable to skip CPU checks for Polars as a fallback
+ENV POLARS_SKIP_CPU_CHECK=1
+

omni-reader/settings.json

@@ -0,0 +1,14 @@
+{
+    "editor.formatOnSave": true,
+    "python.defaultInterpreterPath": "/opt/venv/bin/python",
+    "python.analysis.extraPaths": ["/workspace"],
+    "workbench.startupEditor": "none",
+    "terminal.integrated.defaultProfile.linux": "bash",
+    "workbench.colorTheme": "Dark Modern",
+    "python.linting.enabled": true,
+    "python.linting.pylintEnabled": true,
+    "files.exclude": {
+        "**/__pycache__": true,
+        "**/.pytest_cache": true
+    }
+}