gridoai/gridoai-api

Set up

• Docker
• Docker Compose
• Scala CLI

Development

• Get the .env file
• Start a local Postgres and Redis server using make postgres-up
• (Optional) Start a local gridoai-ml instance for faster and cheaper development
• Run make dev

Deploying

• Set KEY_PATH to the path of your private key file
• Run make deploy

Changing LLM settings

• Start the litellm server using litellm --config litellm-config.yaml. This will print in which port the server is running.
• Set the litellm endpoint in the .env file, e.g. OPENAI_ENDPOINT=http://0.0.0.0:46822

Testing

• Create a test.env file
  • USE_MOCK_LLM should be set to true
  • Redis and Postgres connection strings should be set to the test containers
  • Embedding api endpoint should be set to a local instance of gridoai-ml
• Start the API server using make dev
• Start the test containers using docker compose up
• Run make test to run the tests