Simplify your use of foundation models with the new OCI Data Science AI Quick Actions feature. It’s designed to let anyone easily deploy, fine-tune, and evaluate open source foundation models.
Watch the product tour (6:26)
Kick-start your machine learning projects—without the guesswork. Whether you're exploring ideas or scaling to production, the OCI Data Science AI Quick Actions feature gives you a fast, intuitive way to access, experiment with, and deploy a wide range of open source foundation models.
With built-in MLOps support from OCI Data Science, the AI Quick Actions feature lets you easily import, fine-tune, evaluate, and deploy your models—all in one place.
Choose from a growing catalog of enterprise-ready, open source models tailored to your use cases, expertise, and budget.
Simplify your model deployment using this secure, enterprise-ready workflow. There’s no need to set up a container and go through a complex workflow before deploying your models.
Get started quickly with cached models, including Phi, Falcon, Mistral, Granite, and Llama, for text generation or multimodal tasks. You can also import your models directly from Hugging Face.
Customize models with your own data to improve accuracy for your specific use case. Fine-tuned models are securely saved in your private model repository.
Compare model performance using built-in evaluation tools. Generate detailed reports using metrics such as BLEU, ROUGE, and BERTScore.
Deploy your models and start inferencing. Use the integrated playground to test and iterate quickly.
Meta’s Llama 4 models—Llama 4 Scout and Llama 4 Maverick—are here! These models let you build more personalized multimodal experiences. They offer significant improvements in image and text understanding and instruction following. With OCI Data Science AI Quick Actions supporting the newly released vLLM 0.8.3, you can now deploy and fine-tune Llama 4 models in a no-code environment. This lets more of your team members leverage these models and simplifies the process of working with OCI’s infrastructure and tools.
Read the complete post