strisunshinewentingwang/pytorch

Pull Requests Commits

Use out variants of functions used by linalg.norm, where possible (#45641)

kurtamohler committed 3 years ago

d360402f

Enable CUDA Fuser for ROCm (#45965)

t-vi committed 3 years ago

d3d8da7a

Revert D24099167: [HTE @ clang-tidy] Enable clang-tidy configs inheretence for caffe2 project

suo committed 3 years ago

40828b68

Pin libuv to 1.39 in Windows CI in order to keep version alignment in read me document (#46015)

gunandrose4u committed 3 years ago

283ae199

[StaticRuntime] Replace hashtable based workspace with vector<IValue> (#45892)

Hao Lu committed 3 years ago

ea4fbb2e

Add complex32 dtype support to CPU/GPU implementation of (#45339)

soulitzer committed 3 years ago

735d5b89

Fix doc about operator benchmark (#45853)

ShijunK committed 3 years ago

7d4f5060

[torchscript] Verbose logging of code location causing the error (#45908)

tgalkovskyi committed 3 years ago

acca11b8

unify reproducibility notes (#45748)

ngimel committed 3 years ago

52f2db75

[HTE @ clang-tidy] Enable clang-tidy configs inheretence for caffe2 project

ivanmurashko committed 3 years ago

d93cae00

Fix TypeError when torch.jit.load is passed a pathlib.Path (#45825)

jonathan-conder-sm committed 3 years ago

9dc9a55b

[TensorExpr] LoopNest: add a constructor that takes Stmt instead of list of Tensors. (#45949)

Mikhail Zolotukhin committed 3 years ago

6e4de445

[TensorExpr] LoopNest: replace output_tensors_ with output_bufs_. (#45948)

Mikhail Zolotukhin committed 3 years ago

1036b774

[TensorExpr] Loopnest: unify intermediate_tensors_ and temp_bufs_. (#45947)

Mikhail Zolotukhin committed 3 years ago

29da553d

[TensorExpr] Add shorthand versions for `splitWith{Mask,Tail}` functions. (#45946)

Mikhail Zolotukhin committed 3 years ago

598caddd

[TensorExpr] Nuke `Function` class and directly use `Tensor` instead. (#45936)

Mikhail Zolotukhin committed 3 years ago

b65ffa36

Throw special exception when backend compilation is met with fatal error (#45952)

Yinghai Lu committed 3 years ago

c9caa828

[Onnxifi] Don't throw exception when we cannot write out debug files (#45979)

Yinghai Lu committed 3 years ago

a92b49f7

Run gradgradcheck on torch.fft transforms (#46004)

peterbell10 committed 3 years ago

99d3f37b

Add torch::cuda::ncll::all2all (#45900)

malfet committed 3 years ago

c19b9cd1

Add `tensor_split` function, based on `numpy.array_split` (#45168)

kurtamohler committed 3 years ago

ef4817fe

Workaround for cublas bug for 45724 (#46001)

zasdfgbnm committed 3 years ago

b2bff9e4

codegen: Improve array default handing (#45163)

peterbell10 committed 3 years ago

8d14b50e

[FX] Preserve type annotations on generated code in Graph (#45880)

James Reed committed 3 years ago

00b8ebe6

Add `[zc]heevd` to the list of MKL symbols exported from torch_cpu (#46002)

malfet committed 3 years ago

81d40aaf

Fix cholesky TF32 tests (#45492)

invalid-email-address committed 3 years ago

c59c4b0d

CUDA BFloat16 support of clamp, remainder, lshift, rshift (#45247)

zasdfgbnm committed 3 years ago

903acc6b

Fix distributed documentation for asynchronous collective Work objects (#45709)

rohan-varma committed 3 years ago

154347d8

[NNC] Registerizer V2, supporting partial and conditional replacement (#45574)

nickgg committed 3 years ago

19da1d22

[FakeLowP] T76913842 Make AddFakeFp16 take int inputs (#45992)

venkatacrc committed 3 years ago

a36f11a3

Newer Older