strisunshinewentingwang/pytorch

Pull Requests Commits

Upgrade openmpi for testing

agolynski committed 3 years ago

259fab21

WIP, DO NOT REVIEW enable CUDA aware testing

agolynski committed 3 years ago

0379bdb0

relax test deadlines

ngimel committed 3 years ago

45aa54d8

Expose `cudaMemGetInfo` (#58635)

coreylammie committed 3 years ago

b4b95fc8

[PyTorch] Extract non-template parts of torch::class_ (#54548)

swolchok committed 3 years ago

133133af

.github: Ensure 7zip install for windows (#58924)

seemethere committed 3 years ago

ec89bf65

[Pytorch Delegated Backend] Save function name in debug info (#57481)

kimishpatel committed 3 years ago

ede3f542

[Pytorch Delegated Backend] Save operator name and function name in (#57441)

kimishpatel committed 3 years ago

813adf10

Add no-grad inference mode note (#58513)

soulitzer committed 3 years ago

a7a5992d

Add parsing logic for `Tuple[()]` annotation (#58340)

Ansley Ussery committed 3 years ago

5268b5a2

OpInfo: `diag_embed`, `diagonal` (#58642)

krshrimali committed 3 years ago

b9d1ad9c

Run pthreadpool with _NoPThreadPoolGuard on the same thread (#58759)

Akshit Khurana committed 3 years ago

f9762758

[NNC] Add documentation for splitWith APIs (#58270)

navahgar committed 3 years ago

b703f1e0

[NNC] Make splitWithMask transform in-place (#58269)

navahgar committed 3 years ago

dd7bbe1a

[NNC] Make splitWithTail transform in-place (#58268)

navahgar committed 3 years ago

e2467cc4

[jit] Add Python API for ScriptProfile (#57398)

zhxchen17 committed 3 years ago

6b6a27e4

[resubmit] masked_scatter thrust->cub (#58865)

zasdfgbnm committed 3 years ago

c8833348

Check memory overlap in sort for large input sizes (#58327)

eqy committed 3 years ago

fedf6f2d

[PyTorch] Reduce errors of `foreach` functions (#56993)

crcrpar committed 3 years ago

7eade660

various TensorIterator speed improvements (#58810)

ngimel committed 3 years ago

8a28bbee

Add mish activation function (#58648)

Adnios committed 3 years ago

09a8f22b

Re-enable torchdeploy oss tests and move to per-PR cuda11 job (#58872)

wconstab committed 3 years ago

bf269fdc

[c10d] Use pg wrapper in detailed debug mode (#58281)

rohan-varma committed 3 years ago

19bcbfc5

Disable test_nccl_errors_blocking_abort (#58921)

malfet committed 3 years ago

aad2ad88

[Pytorch] Update FuseLinear to map source range information (#58492)

kimishpatel committed 3 years ago

470160ad

[Pytorch] Provide API to preserve source range and callstack information during graph rewrite (#58300)

kimishpatel committed 3 years ago

e0676751

Increase mimimum number of warmup runs to 2 (#58801)

t-vi committed 3 years ago

2ef9a1df

Forward AD formulas batch 1 (#57768)

albanD committed 3 years ago

09a1b1cf

[torch][repeat_interleave] Fix ambigious function call (#58881)

Serhat Yilmaz committed 3 years ago

b4f3a989

Port elu_backward to structured (#58660)

Freey0 committed 3 years ago

3dbfaddf

Older