mirror of
https://github.com/Memo-2023/mana-monorepo.git
synced 2026-05-15 00:01:10 +02:00
1495dbe476
Python/FastAPI service providing unified OpenAI-compatible API for Ollama and cloud LLM providers (OpenRouter, Groq, Together). Features: - Chat completions with streaming (SSE) - Vision/multimodal support - Embeddings generation - Multi-provider routing (provider/model format) - Prometheus metrics - Optional Redis caching
50 lines
1.2 KiB
YAML
50 lines
1.2 KiB
YAML
version: "3.8"
|
|
|
|
services:
|
|
mana-llm:
|
|
build:
|
|
context: .
|
|
dockerfile: Dockerfile
|
|
container_name: mana-llm
|
|
ports:
|
|
- "3025:3025"
|
|
environment:
|
|
- PORT=3025
|
|
- LOG_LEVEL=info
|
|
- OLLAMA_URL=http://host.docker.internal:11434
|
|
- OLLAMA_DEFAULT_MODEL=gemma3:4b
|
|
- OLLAMA_TIMEOUT=120
|
|
- REDIS_URL=redis://redis:6379
|
|
# Add API keys via .env file
|
|
- OPENROUTER_API_KEY=${OPENROUTER_API_KEY:-}
|
|
- GROQ_API_KEY=${GROQ_API_KEY:-}
|
|
- TOGETHER_API_KEY=${TOGETHER_API_KEY:-}
|
|
- CORS_ORIGINS=http://localhost:5173,http://localhost:3000,https://mana.how
|
|
depends_on:
|
|
- redis
|
|
restart: unless-stopped
|
|
healthcheck:
|
|
test: ["CMD", "python", "-c", "import httpx; httpx.get('http://localhost:3025/health').raise_for_status()"]
|
|
interval: 30s
|
|
timeout: 10s
|
|
retries: 3
|
|
start_period: 10s
|
|
extra_hosts:
|
|
- "host.docker.internal:host-gateway"
|
|
|
|
redis:
|
|
image: redis:7-alpine
|
|
container_name: mana-llm-redis
|
|
ports:
|
|
- "6380:6379"
|
|
volumes:
|
|
- redis-data:/data
|
|
restart: unless-stopped
|
|
healthcheck:
|
|
test: ["CMD", "redis-cli", "ping"]
|
|
interval: 10s
|
|
timeout: 5s
|
|
retries: 5
|
|
|
|
volumes:
|
|
redis-data:
|