STA-AI / config /production.yml
saemstunes's picture
Update config/production.yml
9247f0a verified
environment: production
model:
name: microsoft/Phi-3.5-mini-instruct
quantization: Q4_K_M
context_length: 4096
max_response_length: 500
api:
rate_limiting:
enabled: true
requests_per_minute: 60
burst_capacity: 10
security:
input_sanitization: true
suspicious_pattern_detection: true
max_input_length: 10000
cors:
allowed_origins:
- "https://www.saemstunes.com"
- "https://saemstunes.com"
allowed_methods: ["GET", "POST", "OPTIONS"]
allowed_headers: ["*"]
monitoring:
enabled: true
metrics:
- response_time
- error_rate
- system_resources
- user_activity
- model_performance
alerts:
- high_error_rate:
threshold: 5%
duration: 5m
- slow_response:
threshold: 10s
duration: 2m
- system_overload:
cpu_threshold: 80%
memory_threshold: 90%
duration: 5m
supabase:
max_retries: 3
timeout: 30
connection_pool_size: 10
caching:
enabled: true
ttl: 300
max_size: 1000
logging:
level: INFO
format: json
file_rotation: "10 MB"
max_files: 5
deployment:
primary: "huggingface_spaces"
backup: "railway"
health_check_interval: 30
auto_restart: true