Rename Dockerfile to docker/Dockerfile.prod

Dockerfile

@@ -1,79 +0,0 @@
-# ============================================================================
-# MatAnyOne + SAM2 + Gradio Dockerfile (Python 3.10 + CUDA 12.1)
-# ============================================================================
-
-# Use NVIDIA CUDA 12.1 base image with Python 3.10
-FROM nvidia/cuda:12.1.0-cudnn8-runtime-ubuntu20.04
-
-# Environment settings
-ENV PYTHONDONTWRITEBYTECODE=1 \
-    PYTHONUNBUFFERED=1 \
-    DEBIAN_FRONTEND=noninteractive
-
-# Install Python 3.10 and system dependencies
-RUN apt-get update && apt-get install -y --no-install-recommends \
-    python3.10 python3.10-dev python3-pip python3.10-venv \
-    git wget curl build-essential cmake pkg-config \
-    ffmpeg libsm6 libxext6 libfontconfig1 libxrender1 \
-    libgl1-mesa-glx libglib2.0-0 \
-    libavformat-dev libavcodec-dev libavdevice-dev \
-    libavutil-dev libswscale-dev libswresample-dev libavfilter-dev \
-    && apt-get clean \
-    && rm -rf /var/lib/apt/lists/*
-
-# Create working directory
-WORKDIR /app
-
-# Create non-root user
-RUN useradd -m -u 1000 appuser && chown -R appuser:appuser /app
-USER appuser
-
-# Upgrade pip and install wheel
-RUN python3.10 -m pip install --no-cache-dir --user pip setuptools wheel
-
-# Copy requirements for layer caching
-COPY --chown=appuser:appuser requirements.txt .
-
-# Install PyTorch (CUDA 12.1 build)
-RUN python3.10 -m pip install --no-cache-dir --user \
-    torch==2.1.0 \
-    torchvision==0.16.0 \
-    torchaudio==2.1.0 \
-    --extra-index-url https://download.pytorch.org/whl/cu121
-
-# Install core CV/scientific dependencies
-RUN python3.10 -m pip install --no-cache-dir --user \
-    numpy==1.26.4 \
-    opencv-python-headless==4.10.0.84 \
-    Pillow>=10.0.1,<11.0 \
-    scipy==1.13.1 \
-    av==12.1.0
-
-# Install everything else from requirements.txt
-RUN python3.10 -m pip install --no-cache-dir --user -r requirements.txt
-
-# Add user bin to PATH
-ENV PATH="/home/appuser/.local/bin:$PATH"
-
-# Copy application code
-COPY --chown=appuser:appuser . .
-
-# Create runtime directories
-RUN mkdir -p /tmp/model_cache /tmp/processing /tmp/jobs && \
-    chmod 755 /tmp/model_cache /tmp/processing /tmp/jobs
-
-# Runtime environment variables
-ENV GRADIO_SERVER_NAME="0.0.0.0" \
-    GRADIO_SERVER_PORT="7860" \
-    MODEL_CACHE_DIR="/tmp/model_cache" \
-    TORCH_HOME="/tmp/model_cache"
-
-# Health check
-HEALTHCHECK --interval=30s --timeout=30s --start-period=5s --retries=3 \
-    CMD curl -f http://localhost:7860/ || exit 1
-
-# Expose Gradio port
-EXPOSE 7860
-
-# Run application
-CMD ["python3.10", "app.py"]

docker/Dockerfile.prod

@@ -0,0 +1,47 @@
+version: '3.8'
+
+services:
+  # Main application with GPU support
+  backgroundfx-gpu:
+    build:
+      context: ..
+      dockerfile: docker/Dockerfile
+    image: backgroundfx-pro:gpu
+    container_name: backgroundfx-gpu
+    runtime: nvidia
+    environment:
+      - NVIDIA_VISIBLE_DEVICES=0
+      - CUDA_VISIBLE_DEVICES=0
+      - GRADIO_SERVER_NAME=0.0.0.0
+      - GRADIO_SERVER_PORT=7860
+      - MODEL_CACHE_DIR=/app/models
+      - TORCH_HOME=/app/models/.cache
+      - LOG_LEVEL=INFO
+      - MAX_WORKERS=4
+    volumes:
+      - model-cache:/app/models
+      - uploads:/app/uploads
+      - outputs:/app/outputs
+      - ./config:/app/config:ro
+    ports:
+      - "7860:7860"  # Gradio UI
+      - "8000:8000"  # REST API
+    networks:
+      - backgroundfx-net
+    healthcheck:
+      test: ["CMD", "curl", "-f", "http://localhost:7860/health"]
+      interval: 30s
+      timeout: 10s
+      retries: 3
+      start_period: 60s
+    deploy:
+      resources:
+        reservations:
+          devices:
+            - driver: nvidia
+              count: 1
+              capabilities: [gpu]
+        limits:
+          memory: 16G
+          cpus: '8'
+    restart: