Oracle has announced the general availability of its Oracle Cloud Infrastructure (OCI) Generative AI service, introducing features the company says will aid enterprises in utilizing generative AI advancements. Customers can access the OCI Generative AI service both on the Oracle Cloud and on-premises through OCI Dedicated Region.
The OCI Generative AI service, a fully managed offering, incorporates large language models (LLMs) from Cohere and Meta Llama 2. It comes with multilingual capabilities, supporting over 100 languages. Additionally, the service has improved its GPU cluster management and offers more flexible options for fine-tuning models, says Oracle. Customers may further refine these models using their own data with retrieval augmented generation (RAG) techniques.
The OCI Generative AI service simplifies the customization of generative AI models for text generation, summarization, and semantic similarity tasks, making them accessible via API calls, according to Oracle. Security and data governance are integrated into the technology stack for these models.
Initially supporting OCI OpenSearch, future releases of the OCI Generative AI service will expand to include more data search tools and integration with Oracle Database 23c and MySQL HeatWave. Oracle also plans to incorporate prebuilt agent-actions across its suite of SaaS applications.
Find a blog post by Greg Pavlik, SVP, Oracle Cloud Infrastructure here.
[Image courtesy: Oracle]