strisunshinewentingwang/pytorch

Pull Requests Commits

[dtensor][3/n] use dedicated TensorMeta instead of the fx one (#108261)

wanchaol committed 1 year ago

09f3e08b

[dtensor][2/n] use op overload instead of function schema (#107306)

wanchaol committed 1 year ago

fc1dcfb9

[DCP][Test] Fix device assignment in test/distributed/checkpoint/test_file_system_checkpoint_cpu.py (#109141)

wz337 committed 1 year ago

48e6ffbe

[ONNX] Support None in fx.args as torchlib inputs (#108708)

titaiwangms committed 1 year ago

91e154fc

[HigherOrderOp] Support SymInt as input to body function (#108967)

yanboliang committed 1 year ago

a2ff3454

Update SingletonSymNode to allow more comparisons (#108315)

soulitzer committed 1 year ago

4667a5c9

[pytorch-vulkan] add aten::randn_like & aten::normal_ (#109075)

tina134 committed 1 year ago

a46df6eb

Make mutation test work with quantized tensors (#108935)

ezyang committed 1 year ago

e5f300f0

[submodule] Fix eltwise share buffer issue in ideep (#108038)

yanbing-j committed 1 year ago

687f0278

Add torch.distributed get_rank and get_world_size to constant_fold_functions (#109029)

ezyang committed 1 year ago

e027de2c

Record and replay for ShapeEnv. (#107989)

ysiraichi committed 1 year ago

12e8530b

fix 'Node' object is not iterable in functorch.compile.minifier (#103011)

makslevental committed 1 year ago

e0660564

Add memory overlap check to `meta_copy_` (#108989)

int3 committed 1 year ago

063a6262

Fix cumprod f16 opinfo test via ref-in-float + increasing tolerances (#109128)

int3 committed 1 year ago

f0888528

Fix a bunch of opinfo tests by using reference_in_float (#109089)

int3 committed 1 year ago

6869b25f

Fix std_mean f16 opinfo test by using reference_in_float (#109081)

int3 committed 1 year ago

baefe471

Reland 2: Add PyObject preservation for UntypedStorage (#109039)

kurtamohler committed 1 year ago

4c5e4357

[DTensor] Remove compute_local_offset from _utils.py (#109096)

wz337 committed 1 year ago

6dc56d34

[quant][be] Reduce warnings in tests (#108922)

jerryzh168 committed 1 year ago

cf26e557

assign var for "not populated" str (#108844)

redwrasse committed 1 year ago

9118073f

Revert "[inductor] Lower masked_scatter on CUDA (#108803)"

pytorchmergebot committed 1 year ago

91aab161

[quant][be] Cleanup xnnpack_quantizer implementation (#108921)

jerryzh168 committed 1 year ago

b01b934a

[Gloo] Properly pass op type to Work (#108812)

Rodrigo Kumpera committed 1 year ago

bde75eb9

[Inductor CUTLASS backend] Step 5: Gemm CUTLASS templates (#108015)

ipiszy committed 1 year ago

a2d5f133

[Inductor CUTLASS backend] Step 4: CUDA (template) kernels (#107931)

ipiszy committed 1 year ago

097fd43f

[Inductor CUTLASS backend] Step 3: autotune_process, and CUDABenchmarkRequest (#107901)

ipiszy committed 1 year ago

b2d764ec

[Inductor CUTLASS backend] Step 2: CUDACodeCache (#107847)

ipiszy committed 1 year ago

102fefac

[Inductor CUTLASS backend] Step 1: Inductor config for cuda / cutlass, util functions. (#107802)

ipiszy committed 1 year ago

a14761b6

Revert "CI Sev - pin docker images for A100 workers (#108871)" (#109071)

atalman committed 1 year ago

15b13d3c

make sure all torch._numpy tests run on CI (#108762)

ev-br committed 1 year ago

cd46b5db

Newer Older