Session
Using AI Runtimes for Model Training and Development on Databricks
Overview
| Experience | In Person |
|---|---|
| Type | Breakout |
| Track | Artificial Intelligence |
| Industry | Enterprise Technology |
| Technologies | MLFlow, Mosaic AI |
| Skill Level | Intermediate |
| Duration | 40 min |
Tired of managing complex GPU infrastructure for AI projects? Join us to discover how Databricks' AI Runtime simplifies GPU-accelerated model development with a serverless experience. Learn how the Machine Learning Runtime provides access to pre-configured GPU environments with popular ML frameworks and optimizations for unmatched performance.
Key takeaways:
- Deploying models to serverless GPUs with automatically prepared containers including necessary libraries
- Selecting optimal GPU configurations for different workloads
- Managing costs through automatic scaling based on demand
- Utilizing pre-configured drivers for popular frameworks like TensorFlow and PyTorch
- Achieving 3–5x performance improvements through built-in optimizations
Whether training custom models or fine-tuning foundation models, learn to focus on building AI solutions rather than managing infrastructure.
Session Speakers
IMAGE COMING SOON