ciflow/torchtitan/198536: [DTensor] Migrate unbind sharding propagation to single-dim strategy
- PyTorch: 1453 events in the last 90 days
- PyTorch: 1434th Release in the last 90 days
- Previous: earlier the same day · ciflow/torchtitan/193844: [UPDATE] Update
What happened
aten.unbind.int used a legacy op strategy that enumerated the input's existing sharding candidates and raised RuntimeError whenever the unbind dim was sharded, instead of redistributing. It also recorded an all-zero redistribute cost matrix, understating redistribution cost. Replace it with a single-dim strategy mirroring split: shard rules for every dim except the unbind dim (output placeholder shifted down by one…
Summary assembled by rule from the sources below