Use the seed parameter in OCI Generative AI
- Services: Generative AI
- Release Date: January 30, 2025
OCI Generative AI now supports a seed parameter in the OCI Generative AI playground, CLI, and API. The seed parameter makes a best effort to sample tokens deterministically and the model aims to return the same result for repeated requests when you assign the same seed and parameters for the requests.
This parameter is available for all Meta Llama and Cohere Command models that are supported by the OCI Generative AI service. To use the seed parameter in the API, see CohereChatRequest for Cohere models and GenericChatRequest for Meta Llama models. To use the seed parameter in the playground, see Chat in the Console.
Learn about the definition of the seed parameter and supported foundational models in OCI Generative AI. For information about the service, see the Generative AI documentation.