viable/strict/1789770346: [inductor] Optimize full slice_scatter chains (#195631)
- PyTorch: 1159 events in the last 90 days
- PyTorch: 1145th Release in the last 90 days
- Previous: earlier the same day · viable/strict/1789766664
What happened
The following summary was prepared with assistance from Codex and reviewed by the author. Rewrite slice_scatter chains that fill an entire tensor into peak memory efficient cat or copy_. Functionalizing chunking results in a chain of slice_scatter nodes, holding full-sized intermediate by the end of the chain. The main goal of this pass is to make the "chunking" region the most peak memory efficient, as this is the…
Summary assembled by rule from the sources below