▶ Watch ↗AI Engineer World's Fair 202420:46
Insights from Snorkel AI running Azure AI Infrastructure
Read the full talk →Key ideas
Scroll to read ↓Enterprise model quality depends on the right domain data—and on infrastructure that can turn that data into repeated training, evaluation and synthesis cycles.
- What gets a bank’s model ready to deploy?0:16 ↗
- Scale expert knowledge, then evaluate the actual task3:53 ↗
- Evaluate long context and align models to the domain5:17 ↗
- Optimize the whole infrastructure path6:53 ↗
- PyTorch, Horovod and a shared filesystem8:56 ↗
- Find out why the GPUs are waiting11:05 ↗
- Research needs capacity that can change13:00 ↗
- Compare hardware at approximately equal cost15:10 ↗
- Separate batch-size effects from hardware effects17:02 ↗
- Match evolving accelerators to the workload18:18 ↗
- Turn preference signals and expertise into better data19:46 ↗