Why it matters
Selective recompute improves trade-off. Understanding shapes advanced training.
Advertisement
The architecture
Identify large activations.
Store cheap ones.
Recompute expensive.
~5% compute + significant memory.
Advertisement
How it works end to end
Korthikanti et al. 2022.
Attention scores + dropout expensive to store.
MLP + norm cheap.
~5% compute overhead.