chore: Migrate from Poetry to UV, added health checkpoint, and improved dockerimage

.dockerignore

@@ -0,0 +1,65 @@
+# Version control
+.git
+.gitignore
+.gitattributes
+
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+.Python
+.env
+.venv/
+venv/
+ENV/
+env/
+.pytest_cache/
+.coverage
+htmlcov/
+.tox/
+.nox/
+.hypothesis/
+.mypy_cache/
+
+# IDE
+.idea/
+.vscode/
+*.swp
+*.swo
+.DS_Store
+
+# Project specific
+tests/
+docs/
+*.md
+!README.md
+Makefile
+docker-compose*.yml
+.dockerignore
+Dockerfile*
+
+# Cache and temp files
+*.log
+.cache/
+tmp/
+temp/
+
+# Distribution / packaging
+dist/
+build/
+*.egg-info/
+
+# Local development
+*.local
+.env.local
+.env.*.local
+
+# Model cache and downloads
+model_cache/
+downloads/ 
+
+*.pyc
+*.pyo
+*.pyd
+.env

.gitignore

@@ -18,3 +18,4 @@
 .idea/misc.xml
 .env
 .DS_Store
+__pycache__/

Dockerfile

@@ -1,42 +1,98 @@
-# Use a base image with CUDA support and the desired Python version
-FROM python:3.12-slim-bookworm
-
+FROM ghcr.io/astral-sh/uv:python3.12-bookworm-slim AS builder
 ARG CPU_ONLY=false
+
 WORKDIR /app
 
-RUN apt-get update \
-    && apt-get install -y redis-server libgl1 libglib2.0-0 curl wget git procps \
-    && apt-get clean
+# Install build dependencies
+RUN apt-get update && \
+    apt-get install -y --no-install-recommends libgl1 libglib2.0-0 && \
+    rm -rf /var/lib/apt/lists/*
 
-# Install Poetry and configure it
-RUN pip install poetry \
-    && poetry config virtualenvs.create false
+# Enable bytecode compilation and set proper link mode for cache mounting
+ENV UV_COMPILE_BYTECODE=1 \
+    UV_LINK_MODE=copy \
+    HF_HOME=/app/.cache/huggingface \
+    TORCH_HOME=/app/.cache/torch \
+    PYTHONPATH=/app \
+    OMP_NUM_THREADS=4
 
-COPY pyproject.toml poetry.lock ./
+# Copy dependency files and README
+COPY pyproject.toml uv.lock README.md ./
 
-# Install dependencies before torch
-RUN poetry install --no-interaction --no-root
+# Install dependencies but not the project itself
+RUN --mount=type=cache,target=/root/.cache/uv \
+    uv sync --frozen --no-install-project
+
+# Copy the rest of the project
+COPY . .
 
-# Install PyTorch separately based on CPU_ONLY flag
-RUN if [ "$CPU_ONLY" = "true" ]; then \
-    pip install --no-cache-dir torch torchvision --extra-index-url https://download.pytorch.org/whl/cpu; \
+# Better GPU detection: Check both architecture and if NVIDIA is available
+RUN ARCH=$(uname -m) && \
+    if [ "$CPU_ONLY" = "true" ] || [ "$ARCH" = "aarch64" ] || [ "$ARCH" = "arm64" ] || ! command -v nvidia-smi >/dev/null 2>&1; then \
+    USE_GPU=false; \
     else \
-    pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121; \
+    USE_GPU=true; \
+    fi && \
+    echo "Detected GPU availability: $USE_GPU" && \
+    # For PyTorch installation with architecture detection
+    uv pip uninstall -y torch torchvision torchaudio || true && \
+    if [ "$USE_GPU" = "false" ]; then \
+    # For CPU or ARM architectures or no NVIDIA
+    echo "Installing PyTorch for CPU" && \
+    uv pip install --no-cache-dir torch torchvision --extra-index-url https://download.pytorch.org/whl/cpu; \
+    else \
+    # For x86_64 with GPU support
+    echo "Installing PyTorch with CUDA support" && \
+    uv pip install --no-cache-dir torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121; \
     fi
 
-ENV HF_HOME=/tmp/ \
-    TORCH_HOME=/tmp/ \
-    OMP_NUM_THREADS=4
+# Install the project in non-editable mode
+RUN --mount=type=cache,target=/root/.cache/uv \
+    uv sync --frozen --no-editable
+
+# Download models for the pipeline
+RUN uv run python -c "from docling.pipeline.standard_pdf_pipeline import StandardPdfPipeline; artifacts_path = StandardPdfPipeline.download_models_hf(force=True)"
+
+# Pre-download EasyOCR models with better GPU detection
+RUN ARCH=$(uname -m) && \
+    if [ "$CPU_ONLY" = "true" ] || [ "$ARCH" = "aarch64" ] || [ "$ARCH" = "arm64" ] || ! command -v nvidia-smi >/dev/null 2>&1; then \
+    echo "Downloading EasyOCR models for CPU" && \
+    uv run python -c "import easyocr; reader = easyocr.Reader(['fr', 'de', 'es', 'en', 'it', 'pt'], gpu=False); print('EasyOCR CPU models downloaded successfully')"; \
+    else \
+    echo "Downloading EasyOCR models with GPU support" && \
+    uv run python -c "import easyocr; reader = easyocr.Reader(['fr', 'de', 'es', 'en', 'it', 'pt'], gpu=True); print('EasyOCR GPU models downloaded successfully')"; \
+    fi
 
-RUN python -c 'from docling.pipeline.standard_pdf_pipeline import StandardPdfPipeline; artifacts_path = StandardPdfPipeline.download_models_hf(force=True);'
+# Production stage
+FROM ghcr.io/astral-sh/uv:python3.12-bookworm-slim
+WORKDIR /app
 
-# Pre-download EasyOCR models in compatible groups
-RUN python -c 'import easyocr; \
-    reader = easyocr.Reader(["fr", "de", "es", "en", "it", "pt"], gpu=True); \
-    print("EasyOCR models downloaded successfully")'
+# Install runtime dependencies
+RUN apt-get update && \
+    apt-get install -y --no-install-recommends redis-server libgl1 libglib2.0-0 curl && \
+    rm -rf /var/lib/apt/lists/*
 
-COPY . .
+# Set environment variables
+ENV HF_HOME=/app/.cache/huggingface \
+    TORCH_HOME=/app/.cache/torch \
+    PYTHONPATH=/app \
+    OMP_NUM_THREADS=4 \
+    UV_COMPILE_BYTECODE=1
 
-EXPOSE 8080
+# Create a non-root user
+RUN useradd --create-home app && \
+    mkdir -p /app && \
+    chown -R app:app /app /tmp
 
-CMD ["poetry", "run", "uvicorn", "--port", "8080", "--host", "0.0.0.0", "main:app"]
+# Copy the virtual environment from the builder stage
+COPY --from=builder --chown=app:app /app/.venv /app/.venv
+ENV PATH="/app/.venv/bin:$PATH"
+
+# Copy necessary files for the application
+COPY --chown=app:app . .
+
+# Switch to non-root user
+USER app
+
+EXPOSE 8080
+CMD ["uvicorn", "main:app", "--port", "8080", "--host", "0.0.0.0"]

Makefile

@@ -0,0 +1,125 @@
+# Variables
+PYTHON = poetry run
+DOCKER_CPU_COMPOSE = docker-compose -f docker-compose.cpu.yml
+DOCKER_GPU_COMPOSE = docker-compose -f docker-compose.gpu.yml
+DOCKER_IMAGE = docling-api
+PORT = 8080
+WORKERS = 4
+
+.PHONY: help install dev-setup start stop clean docker-* test lint format
+
+help:
+	@echo "Available commands:"
+	@echo "Development:"
+	@echo "  install         - Install project dependencies using Poetry"
+	@echo "  dev-setup      - Setup development environment (install Redis, etc.)"
+	@echo "  start          - Start all development services locally"
+	@echo "  stop           - Stop all development services"
+	@echo "  clean          - Clean up temporary files and caches"
+	@echo ""
+	@echo "Docker:"
+	@echo "  docker-build-cpu   - Build Docker image (CPU version)"
+	@echo "  docker-build-gpu   - Build Docker image (GPU version)"
+	@echo "  docker-start       - Auto-detect system and start appropriate container (CPU/GPU)"
+	@echo "  docker-start-cpu   - Start services in CPU mode"
+	@echo "  docker-start-gpu   - Start services in GPU mode"
+	@echo "  docker-stop        - Stop all Docker services"
+	@echo "  docker-clean       - Clean Docker resources"
+	@echo ""
+	@echo "Code Quality:"
+	@echo "  format         - Format code using black"
+	@echo "  lint           - Run linter"
+	@echo "  test           - Run tests"
+
+# Development commands
+install:
+	curl -sSL https://install.python-poetry.org | python3 -
+	poetry install
+
+dev-setup:
+	@echo "Setting up development environment..."
+	@if [ "$(shell uname)" = "Darwin" ]; then \
+		brew install redis; \
+		brew services start redis; \
+	elif [ -f /etc/debian_version ]; then \
+		sudo apt-get update && sudo apt-get install -y redis-server; \
+		sudo service redis-server start; \
+	fi
+	@echo "Creating .env file..."
+	@echo "REDIS_HOST=redis://localhost:6379/0" > .env
+	@echo "ENV=development" >> .env
+
+start:
+	@echo "Starting FastAPI server..."
+	$(PYTHON) uvicorn main:app --reload --port $(PORT) & \
+	echo "Starting Celery worker..." && \
+	$(PYTHON) celery -A worker.celery_config worker --pool=solo -n worker_primary --loglevel=info & \
+	echo "Starting Flower dashboard..." && \
+	$(PYTHON) celery -A worker.celery_config flower --port=5555
+
+stop:
+	@echo "Stopping services..."
+	@pkill -f "uvicorn main:app" || true
+	@pkill -f "celery" || true
+	@if [ "$(shell uname)" = "Darwin" ]; then \
+		brew services stop redis; \
+	elif [ -f /etc/debian_version ]; then \
+		sudo service redis-server stop; \
+	fi
+
+# Docker commands
+docker-build-cpu:
+	docker build --build-arg CPU_ONLY=true -t $(DOCKER_IMAGE):cpu .
+
+docker-build-gpu:
+	docker build -t $(DOCKER_IMAGE):gpu .
+
+docker-start-cpu:
+	$(DOCKER_CPU_COMPOSE) up --build --scale celery_worker=1
+
+docker-start-gpu:
+	$(DOCKER_GPU_COMPOSE) up --build --scale celery_worker=3
+
+# Auto-detect architecture and start appropriate container
+docker-start:
+	@echo "Auto-detecting system architecture..."
+	@if [ "$(shell uname -m)" = "arm64" ] || [ "$(shell uname -m)" = "aarch64" ] || ! command -v nvidia-smi >/dev/null 2>&1; then \
+		echo "ARM architecture or NVIDIA drivers not detected. Using CPU mode."; \
+		$(MAKE) docker-start-cpu; \
+	else \
+		echo "NVIDIA GPU detected. Using GPU mode."; \
+		$(MAKE) docker-start-gpu; \
+	fi
+
+docker-stop:
+	$(DOCKER_CPU_COMPOSE) down
+	$(DOCKER_GPU_COMPOSE) down
+
+docker-clean:
+	docker system prune -f
+	docker volume prune -f
+
+# Code quality commands
+format:
+	$(PYTHON) black .
+
+lint:
+	$(PYTHON) flake8 .
+	$(PYTHON) mypy .
+
+test:
+	$(PYTHON) pytest
+
+clean:
+	find . -type d -name "__pycache__" -exec rm -rf {} +
+	find . -type f -name "*.pyc" -delete
+	find . -type f -name "*.pyo" -delete
+	find . -type f -name "*.pyd" -delete
+	find . -type f -name ".coverage" -delete
+	find . -type d -name "*.egg-info" -exec rm -rf {} +
+	find . -type d -name "*.egg" -exec rm -rf {} +
+	find . -type d -name ".pytest_cache" -exec rm -rf {} +
+	find . -type d -name ".mypy_cache" -exec rm -rf {} +
+	find . -type d -name ".tox" -exec rm -rf {} +
+	find . -type d -name "build" -exec rm -rf {} +
+	find . -type d -name "dist" -exec rm -rf {} +