strisunshinewentingwang/pytorch

Pull Requests Commits

registration options should only be callable on rvalues (#21079)

smessmer committed 5 years ago

b91e0d14

Implement various AliasAnalysis operations directly on top of MemoryLocations. (#21203)

Owen Anderson committed 5 years ago

18179217

Cuda persistent softmax (#20827)

thorjohnsen committed 5 years ago

e098878d

Move legacy TH functions(sinh,cosh) to TensorIterator + Vec256 (#21115)

xta0 committed 5 years ago

052bab70

remove quantize_linear from Tensor method (#21196)

jerryzh168 committed 5 years ago

7f960a9c

remove dependency to caffe2::math and eigen (#21169)

jspark1105 committed 5 years ago

c185145d

improve test_docs_coverage error messages (#21029)

Brennan Vincent committed 5 years ago

8c927b20

Fix weak module cuda() _flat_weights bug (#21107)

davidriazati committed 5 years ago

e13b483f

introduce a new intrace to add op [C2 changes] (#21148)

mingzhe09088 committed 5 years ago

0223d374

introduce a new interface to add op [core changes] (#21147)

mingzhe09088 committed 5 years ago

31089b02

Revert D15454048: Move THCTensor_{normal, normal_means, normal_stddevs, normal_means_stddevs} to ATen

ezyang committed 5 years ago

012069ca

Revert D15454052: Move THCTensor_(cauchy) to ATen

ezyang committed 5 years ago

dc8f306b

remove import torchvision when testing torch.hub (#21132)

Ailing Zhang committed 5 years ago

be9ce631

Revert D15558784: [reland][pt1][quant] remove quantize_linear from Tensor method

ezyang committed 5 years ago

e161360b

smessmer committed 5 years ago

5fcd37bd

remove quantize_linear from Tensor method (#21156)

jerryzh168 committed 5 years ago

f91f2476

replace `num_bits` with `quant_min` and `quant_max` (#21097)

jerryzh168 committed 5 years ago

0a0ff831

Add torch.load/torch.save for QTensor (#20830)

jerryzh168 committed 5 years ago

277bf69f

Make CUDA triu / tril support batches of size > 65535 (#21067)

vishwakftw committed 5 years ago

eb4d43df

on import, register class before defining it (#21182)

suo committed 5 years ago

057ddab7

Move THCTensor_(cauchy) to ATen (#20622)

syed-ahmed committed 5 years ago

d6438c95

Move THCTensor_{normal, normal_means, normal_stddevs, normal_means_stddevs} to ATen (#20621)

syed-ahmed committed 5 years ago

26d16ae5

Automatic update of fbcode/onnx to 9005291283e943f1a91da5f0acf218bc4e8eb2ca (#21057)

houseroad committed 5 years ago

07ac00d2

Updated scalar type to onnx mapping (#21095)

izdeby committed 5 years ago

ff0d00f9

Use QTensor for bias (#21038)

dskhudia committed 5 years ago

726caeac

Enable all and any for bool tensors (#21033)

izdeby committed 5 years ago

64f06d49

Enabled add, sum and mul for bool tensor (#21032)

izdeby committed 5 years ago

9a22cb9f

Support for rudimentary f-strings (#21037)

James Reed committed 5 years ago

fe396024

Record source/line info in SourceRange and report in highlight (#21157)

James Reed committed 5 years ago

76deb450

Optimize alias analysis (#20899)

Chillee committed 5 years ago

41635764

Newer Older