strisunshinewentingwang/pytorch

Pull Requests Commits

restore generic IndexToScatterGatherOffset specialization (#40349)

ngimel committed 4 years ago

3bbedb34

Add thrift and tensorpipe backend tests for test_ddp_under_dist_autograd. (#40210)

pritamdamania committed 4 years ago

e632bf8d

Fix update_s3_html for nightly jobs (#40338)

peterjc123 committed 4 years ago

ac8c3c0a

[vulkan] jit passes for vulkan conv2 prepack and fuse with clamp (#39282)

IvanKobzarev committed 4 years ago

38522151

Add unit test to verify DDP + RPC correctness. (#40139)

pritamdamania committed 4 years ago

f69460d0

Change memory format promotion rules of point wise operators. (#37968)

VitalyFedyunin committed 4 years ago

a47fb579

[android][test_app][reland] test_app example linking to pytorch_android aar content (#40313)

IvanKobzarev committed 4 years ago

c1dfc05c

[PyTorch Numeric Suite] Add support for dynamic LSTM (#40065)

Haixin Liu committed 4 years ago

4cbf87dc

Remove incorrect warning message on rounding mode (#40301)

raghuramank100 committed 4 years ago

0079e429

[quant][graphmodel] linear_relu (#40021)

z-a-f committed 4 years ago

9da277c6

[quant][graphmode] clang format changes (#40329)

jerryzh168 committed 4 years ago

e04a611b

[quant][graphmode] docstrings for top level APIs (#40328)

jerryzh168 committed 4 years ago

59ca1d31

[caffe2] optimize 2/4-bit row-wise quantization (#387)

jspark1105 committed 4 years ago

7a837019

Update XLAPreAutograd keys. (#40265)

Ailing Zhang committed 4 years ago

cfe1c6ef

fix small typo in optim adamw (#40283)

lixinyu committed 4 years ago

5c133eb2

[jit] support pad_sequence/pack_sequence (#39844)

wanchaol committed 4 years ago

4b028a8e

Back out "[pytorch][PR] Removes dunder div"

Mike Ruberry committed 4 years ago

4f761f32

Cleanup TensorIteratorDynamicCasting.h (#39839)

zasdfgbnm committed 4 years ago

5555d210

[quant][graphmode] Rename graph mode quantization API to `quantize_jit` (#40212)

jerryzh168 committed 4 years ago

b2f489dc

rename the LayerNorm operator and add it to the replacement map (#40318)

Hector Yuen committed 4 years ago

6d70d157

Make dynamic casting case also benefit from unrolling (#34749)

zasdfgbnm committed 4 years ago

fb17b05f

Add _enable_record_function python API (#40306)

Ilia Cherniavskii committed 4 years ago

41944561

[Resubmit] Ensure NCCL_BLOCKING_WAIT=1 works for dist.barrier() (#40249)

pritamdamania committed 4 years ago

a80dd02a

Add a warning to mention that async_execution does not work with autograd profiler (#40309)

mrshenli committed 4 years ago

314d645e

Minor RPC doc improvements (#40305)

mrshenli committed 4 years ago

5d004438

Fix RRef to_here() docs (#40300)

mrshenli committed 4 years ago

a9f01562

Fix RPC API doc links (#40299)

mrshenli committed 4 years ago

caf0c286

Improve docs for init_rpc (#40298)

mrshenli committed 4 years ago

d6d57939

Improve RPC documents (#40296)

mrshenli committed 4 years ago

3ca05500

Let torch.futures.wait_all re-throw errors (#40291)

mrshenli committed 4 years ago

4463f59c

Newer Older