Accelerate LLM fine-tuning and production deployment
Securely deploy fine-tuned foundation models on any infrastructure with Domino and NVIDIA NIM
Watch on demand
Using generative AI and large language models (LLMs) with proprietary data holds immense potential. Yet the time- and resource-intensive development process makes it difficult to scale these LLMs into production workloads.
While commercial LLM APIs offer convenience, they lack customization, security, and cost-effectiveness. Part of NVIDIA AI Enterprise, NVIDIA NIM inference microservices is a set of containerized inference microservices for accelerating the deployment of foundation models tailored to your specific needs and data — securely and cost-effectively.
In this on-demand webinar, we’ll demonstrate how deploying a NIM model alongside Domino’s Enterprise AI Platform can accelerate your AI development lifecycle through optimized resource utilization, enhanced performance, reduced operational costs, and industry-leading governance.
What you will learn:
- Impact of NVIDIA NIM: Learn how NVIDIA NIM accelerates the pathway for deploying foundation models across various infrastructures – including local workstations, cloud environments, and on-premises data centers.
- Optimized resource utilization and performance: Run a NIM container and Domino in the same cluster for efficient resource allocation and cost savings.
- Robust governance and security: Centralize governance by securely connecting to a NIM foundation model using Domino AI Gateway for controlled user access and comprehensive LLM activity logging.
- Advanced model training and deployment: Use NVIDIA NeMo in Domino to fine-tune a NIM foundation model using parameter efficient fine-tuning (PEFT) - with full reproducibility.
- Scalable, flexible deployment: Deploy a fine-tuned model back to NIM for flexibility to leverage hybrid and multi-cloud environments for resilient AI infrastructure at production scale.
![](https://cdn.sanity.io/images/kuana2sp/production-main/126b264b3af06bf9ea99253e6fe4ecb7defd26d0-1920x675.webp?w=1920&fit=max&auto=format)
![](https://cdn.sanity.io/images/kuana2sp/production-main/f99cbb5933e3c84101c6f244164f5f623123d1c6-1020x220.webp?w=300&fit=max&auto=format)
Featured Speaker
![Josh Mineroff Headshot](https://cdn.sanity.io/images/kuana2sp/production-main/873adef4a890483bc14c59d5148fb3d2ae4922e2-800x800.jpg?w=450&fit=max&auto=format)
Josh Mineroff PhD
Director of Solution Architecture, Tech Alliances, Domino
Josh Mineroff works with partners such as NVIDIA to bring cutting-edge AI solutions from the data and analytics ecosystem to leading data science teams. Before joining Domino in 2019, he founded Meet 'n Treat — a dog socialization platform. During his Ph.D. in applied optimization, he published papers on heart valve design and uncertainty quantification generating over 350 cumulative citations.
![](https://cdn.sanity.io/images/kuana2sp/production-main/cc8e163ed93c388d1afff5ca12d2a2274bd8cee0-1446x1920.jpg?w=450&fit=max&auto=format)
Eda Johnson
Principal Product Architect, NVIDIA
![](https://cdn.sanity.io/images/kuana2sp/production-main/0137ff0067116b896b41bd4f4f33cba1eaf32746-450x120.png?w=200&fit=max&auto=format)
Eda Johnson drives the design and validation of Enterprise AI products at NVIDIA, collaborating with cross-functional teams of software engineers, product managers, and data scientists. She empowers partners to adopt AI Enterprise Products by providing tailored MLOps/LLMOps solutions that seamlessly integrate with their existing offerings.