strisunshinewentingwang/pytorch

Pull Requests Commits

Revert "[C10] PG observability hooks. (#108815)"

pytorchmergebot committed 1 year ago

ff0358b0

[Inductor] Disallow OpOverloadPacket in ir.FallbackKernel (#110567)

SherlockNoMad committed 1 year ago

37a02659

[C10] PG observability hooks. (#108815)

Rodrigo Kumpera committed 1 year ago

0c7a8777

Implement split_with_sizes backward for NT (#110647)

jbschlosser committed 1 year ago

17348b0f

Make unbind() overrideable for NT subclass (#110646)

jbschlosser committed 1 year ago

48240ec6

[sparse] Add i8i8->i32 support for cuSPARSELt (#110499)

jcaip committed 1 year ago

33da6c89

Fix typo under torch/onnx directory (#110697)

kiszk committed 1 year ago

f7ce19d4

[reland] Update singleton int to error when inequality relation is undefined (#110672)

soulitzer committed 1 year ago

69ea214c

Revert "[HigherOrderOp] expose torch.cond (#110293)"

pytorchmergebot committed 1 year ago

576b80d2

Fix clang-tidy warnings in CUDAPluggableAllocator (#110678)

cyyever committed 1 year ago

e75f2e2e

[HigherOrderOp] expose torch.cond (#110293)

ydwu4 committed 1 year ago

601f8728

[quant][pt2][ROCm] follow-up PR 109908 for miopen_batch_norm (#110653)

jeffdaily committed 1 year ago

e8f1f4ed

Doc test non packages (#110568)

albanD committed 1 year ago

c4db6076

Move Docker official builds to Cuda 12.1.1 (#110703)

atalman committed 1 year ago

a3e5ec45

[cpu] remove vec code for ops that do not support complex no (#110280)

khushi-411 committed 1 year ago

261cae79

Fix #110680 (requires_grad typo in decomp) (#110687)

Chillee committed 1 year ago

ceb773b6

perf(optim/dynamo): shortcut `is_sparse` iteration in SGD multi_tensor (#110648)

jon-chuang committed 1 year ago

d776dd04

Revert tl.int1 casting change for ROCm to avoid hangs (#110531)

jataylo committed 1 year ago

96f616a0

Add test_jit_cuda_fuser to ROCM_BLOCKLIST (#110440)

jataylo committed 1 year ago

6b92c367

Add a script to run iOS test app on AWS Device Farm (#110202)

huydhn committed 1 year ago

65afa760

retain_graph=True in compiled_autograd (#110367)

voznesenskym committed 1 year ago

7d98549c

feat(optim): add SGD sparse multitensor to testing path (#110562)

jon-chuang committed 1 year ago

63fe5de8

vmap: decompose real and imag instead of registering batch rule (#110508)

kshitij12345 committed 1 year ago

371d8ba5

Correct outdated Doxygen link (#110654)

kit1980 committed 1 year ago

e8605f6f

Added strict=True to zip in aot_autograd (#110668)

Chillee committed 1 year ago

6d23193a

perf(inductor): improve `Adam` compile times by shortcutting for loops (via `has_complex`) (#110607)

jon-chuang committed 1 year ago

d2799791

Check for both workflow and job names from Dr.CI (#110661)

huydhn committed 1 year ago

26bfb0fc

[CUDA Host Allocator] Add support of CudaHostRegister (#108488)

banitag1 committed 1 year ago

64583c4d

feat(optim): Add adadelta multi_tensor support for complex, with `has_complex` shortcut (#110631)

jon-chuang committed 1 year ago

57e99690

feat(optim): Add `NAdam`support for complex, with `has_complex` shortcut (#110634)

jon-chuang committed 1 year ago

11047be1

Newer Older