strisunshinewentingwang/pytorch

Pull Requests Commits

Disable flaky test_rref_context_debug_info

mrshenli committed 4 years ago

fa666142

Revert D18864774: polish up overloads on free functions

wanchaol committed 4 years ago

73dd8c00

polish up overloads on free functions (#30356)

Elias Ellison committed 4 years ago

44648896

add tests that schemas are valid (#30749)

Elias Ellison committed 4 years ago

a03581b9

Add glue code to collect debug info from all components

mrshenli committed 4 years ago

e9ca13d7

Fix index out of bound error in Engine::ready_queue_size when called before start_threads

mrshenli committed 4 years ago

8a573620

Adding debugging metrics to process group agent

mrshenli committed 4 years ago

a38c9b1a

handle reassignment to inf and nan (#30877)

Elias Ellison committed 4 years ago

82268bf3

add constant prop for immutable types (#30544)

Elias Ellison committed 4 years ago

3eefc06f

rename shouldAnnotate api (#30543)

Elias Ellison committed 4 years ago

648bb501

Proper print for one element tuple (#30853)

jerryzh168 committed 4 years ago

45f0556b

getQParams return a dictionary of qparams (#30859)

jerryzh168 committed 4 years ago

5bf58274

Updating submodules

svcscm committed 4 years ago

fb36f1c3

Fix missing virtual destructor (#30927)

smessmer committed 4 years ago

536481d9

Custom op autograd tests (#30519)

smessmer committed 4 years ago

528fa737

Move Softshrink activation to Aten(CPU+CUDA) (#30229)

XiaobingSuper committed 4 years ago

daef363b

add the worker IDs outside of addSendRpcBackward to ensure they are (#30914)

rohan-varma committed 4 years ago

4f342a61

MultiMarginCriterion: move scalar_check from codegen to code.

gchanan committed 4 years ago

c75bc906

Use universal references and perfect forwarding in Loops.h. (#30466)

Owen Anderson committed 4 years ago

190dac13

call fp16<->fp32 routines in fbgemm from Half2Float and Float2Half operators (#30715)

jspark1105 committed 4 years ago

6848f9ab

Add debug info API for distributed autograd. (#30642)

pritamdamania committed 4 years ago

776fdda7

Updating submodules

svcscm committed 4 years ago

0b330809

MultiheadAttention fixes

pbelevich committed 4 years ago

4bb497b3

Updating submodules

svcscm committed 4 years ago

8b6d7698

Fix lint issues in dist_autograd_test.py (#30928)

pritamdamania committed 4 years ago

f1bd8cc2

Support exporting aten::copy_ and aten::index_put to ONNX opset 11 (#26941)

BowenBao committed 4 years ago

63f1b780

Enable using `torch.autograd.profiler.record_function` as decorator (#30861)

bddppq committed 4 years ago

a26238da

Attach autograd edges only for tensors requiring grad. (#30904)

pritamdamania committed 4 years ago

5c569867

Actually make flake8 do something (#30892)

suo committed 4 years ago

62b10721

embedding_bag make_bag_size optimization (#30701)

ajyu committed 4 years ago

8d35b6ce

Newer Older