strisunshinewentingwang/pytorch

Pull Requests Commits

[easy] Remove large LayerNorm sample input causing OOM from ModuleInfo (#98424)

mikaylagawarecki committed 1 year ago

3a0ad3c1

Add more debug logs to evaluate_expr (#98344)

ezyang committed 1 year ago

3ed66f94

remove //c10:headers (#98420)

mikey dagitses committed 1 year ago

f557402e

Make the Index Rounding Mode Consistent Between the 2D and 3D GridSample Nearest Neighbor Interpolations (#97000)

leimao committed 1 year ago

937ba248

[dtensor] add placement strategy and einsum strategy (#98227)

wanchaol committed 1 year ago

dcec2100

[pruning][core][feature] Implement convert for pruner (#97545)

jcaip committed 1 year ago

93063768

Add forward_prefetch flag to fully_shard (#98277)

rohan-varma committed 1 year ago

3657b37d

Clarify `Tensor.is_sparse` doc (#98408)

malfet committed 1 year ago

49c13025

Change daily aggregates upload job to use sum and occurence counter instead of averages (#98359)

PaliC committed 1 year ago

d1de5f5f

[spmd compile api] pre-flatten state container and pass the flattened state container to transforms (#98392)

yifuwang committed 1 year ago

762a81cb

[spmd compile API] add a (temporary) mechanism for overriding input tensors' placements (#98391)

yifuwang committed 1 year ago

37dbd5bf

[spmd expansion] support scalar_tensor (#98390)

yifuwang committed 1 year ago

970c08f9

[spmd expansion] speed up expansion by ~5x (#98389)

yifuwang committed 1 year ago

0830808d

[spmd expansion] support torch.ops.aten.sym_numel (#98388)

yifuwang committed 1 year ago

161f7c0b

Pattern matcher improvements (#97740)

jansel committed 1 year ago

3344d79e

Revert "[CUDA12] set_device change (#94864)"

pytorchmergebot committed 1 year ago

279ca5f9

Better Handling of Storage Cache (#98254)

eellison committed 1 year ago

981f9f04

Make sure we dealloc on recording, not just replay (#97440)

eellison committed 1 year ago

f1b901b0

[CUDA12] set_device change (#94864)

Aidyn-A committed 1 year ago

c18be2b2

Fix GridSample Activation Quantization (#98278)

leimao committed 1 year ago

7b088890

Add test for pickle_module (#98373)

malfet committed 1 year ago

3da7e832

add new() method identifier to _StorageBase (#98201)

XDaoHong committed 1 year ago

ea00f850

Only print guard code when printing guards (#98347)

ezyang committed 1 year ago

2a32bc50

Add itemsize and nbytes properties to Tensor (#98322)

bjhargrave committed 1 year ago

555ab310

fix _slice_meta's shape calculation (#98326)

ar0ck committed 1 year ago

14ccad73

Fix complex variable notation for division operator to be consistent. (#98057)

redwrasse committed 1 year ago

b4420f0f

Uniformly use elem when checking ListType (#97873)

dilililiwhy committed 1 year ago

526b564f

[CI] Mark sebotnet33ts_256 as nondeterministic (#98356)

desertfire committed 1 year ago

c4de7fde

[inductor] Fix slow tests not being run in CI (#97841)

peterbell10 committed 1 year ago

a05d787e

Revert "remove typed StorageImpl::data() and StorageImpl::unsafe_data() (#98219)"

pytorchmergebot committed 1 year ago

45edc58e

Newer Older