strisunshinewentingwang/pytorch

Pull Requests Commits

Fix incorrect DDP test (#64074)

rohan-varma committed 3 years ago

59fcbd17

[c10d] Prefer use of torch_check (#63928)

rohan-varma committed 3 years ago

9b8f9d5a

Add fast path for addmm when the inputs are conjugate (#59380)

anjali411 committed 3 years ago

5d80a48c

[DDP Comm Hook] Add bf16 gradient compression to ddp_comm_hooks.rst (#64346)

Yi Wang committed 3 years ago

a8f9aab8

[quant][graphmode][fx] Add fbgemm backend_config_dict (#64288)

jerryzh168 committed 3 years ago

ed89937d

Make datasets in `ConcatDataset` not need to be sized (#64114)

bryant1410 committed 3 years ago

69f4401b

Restore LayerNorm numerics test (#64385)

zou3519 committed 3 years ago

535526b9

[quant][graphmode][api] Add backend_config_dict to prepare_fx api (#64135)

jerryzh168 committed 3 years ago

7ffcf155

Silent rm error for sccache log file (#64388)

zhouzhuojie committed 3 years ago

93bc0362

[xplat][metal] Add getters and setters for ivars in Conv2dOpContext (#57395)

husthyc committed 3 years ago

94956749

[structured] Preserve computed elements from meta func to impl (#61746)

Meghan Lele committed 3 years ago

968d7ee4

[Static Runtime] Make per-op latency readable by FAI-PEP (#64315)

Mike Iovine committed 3 years ago

4aad3661

Update optimize_for_mobile to preserve node's debug information (#63106)

salilsdesai committed 3 years ago

86c96542

Break up "@generated" string so Phabricator shows changes

dreiss committed 3 years ago

15ff25d1

Add forward AD support for custom Functions (#64061)

albanD committed 3 years ago

e322547f

Fix bytes_written and bytes_read (#64244)

motanvfb committed 3 years ago

25e25789

[Caffe2] Create fewer strings during argument fetching (#64285)

swolchok committed 3 years ago

03a58a2b

Back out "Revert D30327514: [Pytorch lite predictor] Use KinetoEdgeCPUProfiler for operator profiling." (#64307)

kimishpatel committed 3 years ago

46800160

Add a record scope around autograd::engine::evaluate_function (#63619)

rohan-varma committed 3 years ago

421d8f86

[Bootcamp] Include both python unittest and parser parameters in --help and -h flag (#64297)

PatrickKan committed 3 years ago

0b48d968

[FX] Fix python code generation for wrapped getattr() with default value (#64271)

yqhu committed 3 years ago

c6505cc3

[nnc] Updated generic error message with info about turning off the fuser (#64316)

navahgar committed 3 years ago

87d8ab6e

Fixes reduction launch config (#64304)

zasdfgbnm committed 3 years ago

c4f3f6e6

OpInfo for `nn.functional.layer_norm` (#63276)

krshrimali committed 3 years ago

d5bfdd3d

fix GradBucket.is_last() logic (#63768)

nimaelyasi committed 3 years ago

d1f3d85f

Revert D29699456: [pytorch][PR] Enable Half, BFloat16, and Complex dtypes for coo-coo sparse matmul [CUDA]

zou3519 committed 3 years ago

92b31b59

[FX] Rename reduce functions back to their old, public names (#64324)

James Reed committed 3 years ago

0c4e4e58

[Metal][GPU] Enable metal for simulators and fix test failures if possible (#64322)

husthyc committed 3 years ago

05ecaefb

[CUDA graphs] hotfix for test_graph_ (#64339)

mcarilli committed 3 years ago

24e50b84

Remove outdated warning about RecursiveScriptModule not being copiable (#64085)

gmagogsfm committed 3 years ago

479fc4e4

Newer Older