Paul Zhang

Fix Pyre check

1

Differential Revision: D59124237

Make PositionWeightedModuleCollection faster with torch.export

1

Differential Revision: D58830349

Reland preserve device in empty lookup case

1

Differential Revision: D58828898

Formatting

1

Summary: arc lint external OSS PR Differential Revision: D57507551

Remove extremely slow logic from RequestOnlySplitEmbeddingBagCollectionsSparseArch

2

Differential Revision: D57290467

Enable TW pruning, no remapping

1

Summary: Enable TW pruning for TorchRec inference modules. We switch from pruned_indices_remapping to num_rows_post_pruning given the fact that the indices remapping isn't calculated until after physical transformations. Logical transformations (optimizing...

Fix torch.export stuckness with APS full model export

2

Summary: Resolve torch.export stuckness in exporting full model IG CTR for PT2 full stack inference workstream: https://docs.google.com/document/d/14LUIir04SkiagOchVpiOrXg-Pqqg5MiG9ArYNzWwyik/edit Differential Revision: D61560171

Make inference discoverable as Python module

3

Differential Revision: D61541879

Remove test_compound_module from OSS

4

Differential Revision: D61151900

Fix GPU CI

2

Differential Revision: D61140212