Level 12: Optimize training speed

In this level you’ll use compilers, advanced profilers and mixed precision techniques to train bigger models faster.

---

Speed up models by compiling them

Use torch.compile to speed up models on modern hardware

intermediate

Explore advanced mixed precision settings

Enable state-of-the-art scaling with advanced mix-precision settings.

intermediate

Enable advanced profilers

Tune model performance with profilers.

intermediate

Profile PyTorch operations

Learn to find bottlenecks in PyTorch operations.

intermediate