strisunshinewentingwang/pytorch

Pull Requests Commits

[export] add replace_set_grad_with_hop_pass (#119810)

ydwu4 committed 218 days ago

812f05d7

[export] add node_inline_ to prepare replacing set_grad_enabled with hop (#119736)

ydwu4 committed 218 days ago

4769e691

[export] add sequential_split to prepare replacing set_grad_enabled with hop (#119732)

ydwu4 committed 218 days ago

068659dd

tiny improvement to the cprofile wrapper (#120100)

shunting314 committed 218 days ago

becfda00

[inductor] logging meta data for inductor generated triton kernel (#120048)

shunting314 committed 218 days ago

36e118b8

Add quantized gelu (#119935)

andre-meta committed 218 days ago

24968ff0

[Memory Snapshot] Add CUDAAllocatorConfig details into snapshot metadata (#119404)

aaronenyeshi committed 218 days ago

7973ac58

[BE] Delete `C10_IS_TRIVIALLY_COPYABLE` (#120120)

malfet committed 218 days ago

9aa8bbf7

Add hpu device support in storage/resize (#119761)

jeejakp12 committed 218 days ago

79569d11

[ONNX][dynamo_export] Adjust to new symbolic shape name format in value_info (#119855)

BowenBao committed 218 days ago

6b63d3ba

Simplify c10::is_pod implementation and remove unneeded inclusion of C++17.h (#118212)

cyyever committed 218 days ago

e61c8ef3

[structural binding][4/N] Replace std::tie with structural binding (#120039)

cyyever committed 218 days ago

6952d6dd

Add FakeTensor support to torch._utils._rebuild_tensor (#108186)

Thiago Crepaldi committed 218 days ago

761fa5d6

Remove unused import (#120004)

tugsbayasgalan committed 218 days ago

7ad4ab47

[PT2][Optimus][Observability] Log the optimus graph transformation to the scuba (#119745)

mengluy0125 committed 218 days ago

7b1f5c87

[dynamo][functional_collectives] Add all_to_all_single, all_gather_list, reduce_scatter_list to dynamo remapping (#119683)

IvanKobzarev committed 218 days ago

006eead7

[Dynamo] Fix ListIteratorVariable repr to avoid log flooding (#120053)

yanboliang committed 218 days ago

4f4629d5

DTensor: make tensor_flatten more compatible for dynamo getattr (#118209)

bdhirsh committed 218 days ago

26343451

dynamo: support attribute access on tensor subclasses without sources (#117666)

bdhirsh committed 218 days ago

ee7bcf23

dynamo: respect autograd.Function + multiple save_for_backward calls (#117667)

bdhirsh committed 218 days ago

67f6aca0

Support broadcast in native funcol (#119229)

yifuwang committed 218 days ago

4ac857f9

[EZ] Fix argument parsing in build_with_debinfo (#120088)

malfet committed 218 days ago

24d5caba

Fix searchsorted function signature in docs (#120086)

malfet committed 218 days ago

2d4aa91a

[Optim][Rprop] Replace new().resize_as_() by torch.full_like() (#119978)

wz337 committed 218 days ago

288d1f36

[quant][pt2e] Add `model_is_exported` util function (#119726)

andrewor14 committed 218 days ago

6ea44808

Rename singleton int to nested int (#119661)

soulitzer committed 219 days ago

312ce35c

Make roll a decomposition and remove its lowering (#119857)

lezcano committed 219 days ago

b97fa6ac

Correct index propagation for % (#119864)

lezcano committed 219 days ago

8b02d641

Simplify indexing when doing ModularIndexing + index propagation. (#119863)

lezcano committed 219 days ago

00524970

Revert "Add pixel_shuffle to core aten decomps (#119899)"

pytorchmergebot committed 219 days ago

86dedebe

Newer Older