Introduction
SSP Cloud is a data processing platform dedicated to experimenting with data science methods using open data. It offers a shared environment for users to launch, customize, and manage various services autonomously. Among the available services, the on-demand provisioning of Large Language Models (LLM) represents a significant advancement. These models enable the automation of complex tasks, accelerate information access, and optimize business processes. LLMs facilitate interactions through natural language processing, thus offering solutions tailored to users’ specific needs. By integrating these technologies, SSP Cloud positions itself as a key lever to improve training and prototyping in the field of artificial intelligence.
Why use SSP Cloud for Training?
The service for providing Large Language Models (LLM) on demand in the SSP Cloud is based on an innovative and modular architecture. At the heart of this architecture is an instance of OpenWebUI, which serves as an open and flexible user interface. This OpenWebUI instance interacts directly with Ollama, an advanced platform for managing and deploying language models. Thanks to this configuration, users can easily access, customize, and deploy LLM models according to their specific needs. The list of available models is managed by the platform administrators and is likely to evolve over time to meet new requirements and technological advancements.
SSPCloud does not provide formal service level agreements (SLAs) or guarantees regarding service availability. This means that users should be prepared for potential service interruptions or downtimes.