server/requirements-advanced.txt

# Advanced ML & Performance Dependencies
# Production-ready, lightweight, optimized

# ===== Core ML/NLP =====
torch>=2.1.0
transformers>=4.35.0
sentence-transformers>=2.2.0  # Sentence embeddings (80MB, fast)
tokenizers>=0.15.0  # Rust-based tokenizer (much faster)
datasets>=2.14.0

# ===== ONNX Optimization (2x faster inference) =====
onnxruntime>=1.16.0  # CPU-optimized inference
optimum>=1.14.0  # Convert models to ONNX

# ===== Sentiment & NER Models =====
# Note: Models auto-downloaded on first use
# - cardiffnlp/twitter-roberta-base-sentiment-latest (~500MB)
# - j-hartmann/emotion-english-distilroberta-base (~250MB)
# - dslim/bert-base-NER (~420MB)
# Total: ~1.2GB one-time download, ~300MB in memory

# ===== Scientific Computing =====
numpy>=1.24.0
scikit-learn>=1.3.0  # For clustering (DBSCAN)
scipy>=1.11.0

# ===== Infrastructure =====
redis>=5.0.0  # Caching and rate limiting
celery>=5.3.0  # Background tasks
kafka-python>=2.0.2  # Event streaming (optional)

# ===== API Framework =====
fastapi>=0.104.0
uvicorn[standard]>=0.24.0
python-multipart>=0.0.6
sse-starlette>=1.8.0  # Server-Sent Events
websockets>=12.0  # WebSocket support

# ===== Database =====
sqlalchemy>=2.0.0
alembic>=1.12.0
psycopg2-binary>=2.9.0  # PostgreSQL
clickhouse-driver>=0.2.6  # Analytics (optional)

# ===== Monitoring & Logging =====
prometheus-client>=0.19.0
python-json-logger>=2.0.0

# ===== Utilities =====
python-dotenv>=1.0.0
pydantic>=2.5.0
httpx>=0.25.0  # Async HTTP client
aiofiles>=23.2.1  # Async file I/O

# ===== Email =====
aiosmtplib>=3.0.0  # Async SMTP

# ===== Testing (Dev) =====
pytest>=7.4.0
pytest-asyncio>=0.21.0
pytest-cov>=4.1.0
httpx>=0.25.0  # For testing FastAPI
faker>=20.0.0  # Generate test data

# ===== Code Quality (Dev) =====
black>=23.11.0
isort>=5.12.0
flake8>=6.1.0
mypy>=1.7.0

# ===== Already Installed (from earlier) =====
# google-generativeai  # Gemini API
# These are already in your requirements.txt, keeping them