Exploring Infrastructure Management for GenAI Beyond Kubernetes
During this talk, we will discuss the Kubernetes stack's drawbacks in the context of AI and show how dstack addresses them in the training and deployment of GenAI models.
Enterprises, inspired by OpenAI and recent advancements in the open-source AI community, are exploring training and deploying GenAI models on their infrastructure. The traditional Kubernetes stack is not tailored for the AI era and presents numerous drawbacks hindering enterprises from adopting GenAI.
In this webinar, we’ll discuss:
- The significance of open source for infrastructure management
- Drawbacks of the Kubernetes stack in the context of AI
- Introducing dstack, a new open-source orchestration engine
- Exploring how dstack addresses the drawbacks of Kubernetes in the training and deployment of GenAI models
- Highlighting dstack’s role in simplifying and optimizing infrastructure management processes
Speaker:
Andrey is a core contributor to dstack. He is passionate about open-source, developer tools, and the cloud. Previously, Andrey worked at JetBrains with the PyCharm team.
Links:
If you want to explore the topic further, here are some resources to start with the dstack and deepen your understanding.
- Getting Started with dstack: Learn how to define dev environments, tasks, and services using any cloud provider.
- Real-life Examples with dstack: Deploying and fine-tuning LLMs and SDXL, building RAGs, and more.
- dstack on GitHub: Want to support dstack on GitHub or become an open-source contributor? Go for it!
And hey, to stay in the loop with the dstack community, why not follow dstack on Twitter?
