strisunshinewentingwang/pytorch

Pull Requests Commits

[FSDP][state_dict] Restore the state_dict_config for NO_SHARD (#100855)

fegin committed 1 year ago

55844dfd

fix _privateuse1_tag problem (#100632)

XDaoHong committed 1 year ago

a723f1f2

[opinfo prims] equal (#100663)

khushi-411 committed 1 year ago

5a933d04

[CUDA][CUBLAS] Fix BF16 reduced precision reduction note in docs (#101044)

eqy committed 1 year ago

33f3dca6

[CUDA][CUBLAS] Add cuBLAS workspace allocation behavior to docs (#100919)

eqy committed 1 year ago

6e2efd16

Apply static policy correctly to unspec (#98983)

ezyang committed 1 year ago

1e89a56a

Fix typo in RELEASE.md and README.md (#100536)

Djacon committed 1 year ago

dfa95117

inductor(CPU): fix issue when padding/stride/dilation size is one for cpu weight packing pass (#100951)

XiaobingSuper committed 1 year ago

2b250e19

PyTorch -> C++17 (#98209) (#100557)

r-barnes committed 1 year ago

083f88e1

[MPS] Fix build regressions introduced by #92868 (#101036)

malfet committed 1 year ago

30cecc0e

[inductor] default cache_dir in torch._inductor.codecache should be lazily evaluated (#100824)

hsfzxjy committed 1 year ago

b004c0b3

CUBLAS Flag (`CUBLAS_GEMM_DFALT_TENSOR_OP` -> `CUBLAS_GEMM_DEFAULT_TENSOR_OP`) (#100976)

DyeKuu committed 1 year ago

b06c180a

[vision hash update] update the pinned vision hash (#101032)

pytorchmergebot committed 1 year ago

535368f0

[c10d] make ProcessGroupNCCL work.wait() respect timeout (#100162)

cdzhan committed 1 year ago

649e6096

[inductor] support vec type conversion between float and bool (#100950)

jgong5 committed 1 year ago

b33c9c7c

Extend assert statement to include ListVariable (#100841)

Bearnardd committed 1 year ago

44e73da4

STFT: correct stft definition and better document tensor shapes (#100427)

peterbell10 committed 1 year ago

27d5019e

Revert "[quant][pt2e] Change input act annotation to a map and allow dynamic quantization for non zeroth argument (#101005)"

pytorchmergebot committed 1 year ago

2241aaa6

[CI] Delete skips from https://github.com/pytorch/pytorch/issues/93847 (#96049)

desertfire committed 1 year ago

76cc3ab4

explicitly check or discard cudaGetLastError return value (#100488)

jeffdaily committed 1 year ago

bf214f40

[quant][pt2e] Change input act annotation to a map and allow dynamic quantization for non zeroth argument (#101005)

jerryzh168 committed 1 year ago

f08ddae8

[BE] Prevent pytest from thinking this class defines any tests (#100949)

ZainRizvi committed 1 year ago

20a231b5

Enable reordering pass (#100747)

eellison committed 1 year ago

6308563a

[dynamo 3.11] fix segfault when printing stack trace (#99934)

williamwen42 committed 1 year ago

7da8705f

Revert "fix specify_constraints's signature when exporting model (#100739)"

pytorchmergebot committed 1 year ago

d57544d3

Revert "[Dynamo] Fix nested function resume execution (#100426)"

pytorchmergebot committed 1 year ago

4b8127b9

[ET selective build] add kernel metadata section to selective_build.yaml (#100665)

kirklandsign committed 1 year ago

fa6df34d

Run C++ tests on CI with run_test.py (#99956)

huydhn committed 1 year ago

35834a40

Add CUTLASS-based MM for structured sparse linear operator (#100485)

alexsamardzic committed 1 year ago

a8c2cd10

[optim] More cleanup and reorg of test_optim.py (#100917)

janeyx99 committed 1 year ago

d63e0b15

Newer Older