ciflow/trunk/187952: Fix out= variant recompilations for torch.max/min/topk
- PyTorch: 248 events in the last 90 days
- PyTorch: 245th Release in the last 90 days
- Previous: earlier the same day · trunk/cbb5d2e8abc050fabcda61dcc95cb21d42fb4691: [torchtitan hash update] update the pinned torchtitan hash (#194670)
What happened
torch.max(x, dim, out=...), torch.min(..., out=...) and torch.topk(..., out=...) recompiled under torch.compile on every new input shape, while the functional variants specialize only once. With the default recompile_limit this also makes the function fall back to eager after 8 recompiles, losing compilation entirely. Root cause: the out overloads aten.max.dim_max, aten.min.dim_min and aten.topk.values had no Python…
Summary assembled by rule from the sources below