Create Dockerfile

Dockerfile

@@ -0,0 +1,41 @@
+# 1. Base Image: Use a slim Python image
+FROM python:3.10-slim
+
+# 2. Set Environment Variables (Optional but good practice)
+ENV PYTHONUNBUFFERED=1 \
+    PIP_NO_CACHE_DIR=off \
+    PIP_DISABLE_PIP_VERSION_CHECK=on \
+    # Set path for Hugging Face model cache (inside the container)
+    HF_HOME="/app/huggingface_cache" \
+    TRANSFORMERS_CACHE="/app/huggingface_cache/transformers" \
+    # If you need to use a Hugging Face token (e.g., for gated models, though Gemma isn't)
+    # HUGGING_FACE_HUB_TOKEN="your_hf_token_here" # Best to pass this at runtime or via secrets
+
+# 3. Set Working Directory
+WORKDIR /app
+
+# 4. Copy requirements file and install dependencies
+# This is done before copying the rest of the app to leverage Docker layer caching.
+COPY requirements.txt .
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    build-essential \
+    && rm -rf /var/lib/apt/lists/* # build-essential might be needed by some torch/scipy deps
+RUN pip install --no-cache-dir -r requirements.txt
+
+# 5. Copy the rest of the application code
+COPY app.py .
+
+# 6. Create the cache directory and set permissions (if needed)
+# The user running the pip install command will own the HF_HOME path.
+# If running as non-root later, ensure permissions.
+RUN mkdir -p $HF_HOME && chmod -R 777 $HF_HOME
+# Note: chmod 777 is very permissive, for production, use a more specific user/group.
+# For Hugging Face Spaces, this is generally fine.
+
+# 7. Expose the port the app runs on
+EXPOSE 8000
+
+# 8. Command to run the application
+# The model will be downloaded on the first run if not already cached.
+# For Hugging Face Spaces, this download happens during the build process.
+CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "8000"]