strisunshinewentingwang/pytorch

Pull Requests Commits

Implement torch.linalg.qr (#47764)

antocuni committed 3 years ago

361f5ed9

fx quant: split linear test cases (#49740)

vkuzo committed 3 years ago

bc4ff7ba

fx quant: hook up ConvTranspose{n}d (#49717)

vkuzo committed 3 years ago

ea558b21

[JIT] Constant prop getattr (#49806)

Elias Ellison committed 3 years ago

fc559bd6

[NNC] masked fill (#49627)

Elias Ellison committed 3 years ago

268441c7

Support the `in` operator with str (#47057)

Ansley Ussery committed 3 years ago

58fe6796

[*.py] Rename "Arguments:" to "Args:" (#49736)

SamuelMarks committed 3 years ago

e6779d43

early termination of CUDA tests (#49869)

Rong Rong (AI Infra) committed 3 years ago

9c64b9ff

[numpy] `torch.digamma` : promote integer inputs to float (#48302)

kshitij12345 committed 3 years ago

963f7629

Revert D25684692: [quant][graphmode][fx] Standalone module support {input/output}_quantized_idxs

Mike Ruberry committed 3 years ago

46cf6d33

Clip small scales to fp16 min

csummersea committed 3 years ago

ec6de6a6

[quant][graphmode][fx] Standalone module support {input/output}_quantized_idxs (#49754)

jerryzh168 committed 3 years ago

89b4899e

Revert D25692616: [pytorch][PR] [reland] Early terminate when CUDA assert were thrown

Rong Rong (AI Infra) committed 3 years ago

69b13735

Creation of test framework for Sparse Operators (#48488)

hlahkar committed 3 years ago

9552cc65

Revert D25690129: [pytorch][PR] Added linalg.inv

Mike Ruberry committed 3 years ago

5acc27c0

Fix typo in complex autograd docs (#49755)

soulitzer committed 3 years ago

18330092

[reland] Early terminate when CUDA assert were thrown (#49799)

Rong Rong (AI Infra) committed 3 years ago

e6a21559

[numpy] `torch.erfinv`: promote integer inputs to float (#49155)

kshitij12345 committed 3 years ago

3f4b98d5

[pt][quant] Make the CUDA fake quantize logic consistent with CPU fake quantize logic (#49808)

jianyuh committed 3 years ago

4d611093

removes more unused THC functions (#49788)

ngimel committed 3 years ago

e1631729

Improve docs for scatter and gather functions (#49679)

rgommers committed 3 years ago

d99a0c3b

Mod lists to neutral+descriptive terms in caffe2/docs (#49803)

r-barnes committed 3 years ago

b3387139

Added linalg.inv (#48261)

IvanYashchuk committed 3 years ago

8554b58f

Preserve memory format in qconv op (#49533)

Akshit Khurana committed 3 years ago

370350c7

[lint doc] how to fix flake errors if pre-commit hook wasn't there (#49345)

stas00 committed 3 years ago

5171bd94

[Gradient Compression] Directly let world_size = group_to_use.size() (#49715)

Yi Wang committed 3 years ago

55b431b1

[Gradient Compression] Explicitly restrict the scope of torch.cuda.synchronize to the current device (#49711)

Yi Wang committed 3 years ago

88c33ff8

torch.utils.checkpoint.checkpoint + torch.cuda.amp (#49757)

mcarilli committed 3 years ago

ee271047

[quant][graphmode][fx] Change standalone module api (#49719)

jerryzh168 committed 3 years ago

f474ffa1

[Gradient Compression] Change wait() to value() in some callbacks of PowerSGD communication hook (#49709)

Yi Wang committed 3 years ago

af1b636b

Newer Older