1 dépôt
Tools for measuring GPU memory consumed by intermediate activations during training, which scales quadratically with sequence length.
Distinct from Training Memory Optimizers: Distinct from Training Memory Optimizers: focuses on profiling and estimating activation memory usage, not on optimization techniques.
Explore 1 awesome GitHub repository matching data & databases · Activation Memory Profilers. Refine with filters or upvote what's useful.
Provides activation memory estimation tools critical for understanding training memory bottlenecks.