6 docs tagged with "service"

The docker service maintains AI model containers on a node. It expects the docker socket to be mounted.

The download service keeps a network local copy of files frequently accessed by services in the Singulatron platform.

The dynamic service is designed to help build backendless applications: the goal is to be able to save and query data directly from the frontend. Similarly to Firebase.

The model service can start, stop AI models across multiple runtimes (eg. Docker) and maintains a database of available models on the platform.

The prompt service provides an easy to use interface to prompt LLMs and use AI models. Aims to serve humans and machines alike with its resilient queue based architecture.

The user service is at the heart of Singulatron, managing users, tokens, organizations, permissions and more. Each service and human on the Singulatron network has an account in the User Svc.