strisunshinewentingwang/pytorch

Pull Requests Commits

[FSDP] Add initial `summon_full_params(with_grads=True)` (#85738)

awgu committed 2 years ago

a95889ba

[optim] fix: empty grad support for SparseAdam (#86459)

kshitij12345 committed 2 years ago

82229d1e

Revert "Disable mac m1 jobs (#86463)"

pytorchmergebot committed 2 years ago

66d480d3

Disable mac m1 jobs (#86463)

huydhn committed 2 years ago

ac632b43

[ao] fixing public v private for fuser_method_mappings.py (#86029)

HDCharles committed 2 years ago

ac74976a

[FSDP] Add `use_orig_params` (#84911)

awgu committed 2 years ago

be682bef

Add reference counter in FileStore (#85601)

KinglittleQ committed 2 years ago

b43ae1c4

[quant][ao_migration] nn.intrinsic.qat migration to ao (#86171)

zaf committed 2 years ago

efccb640

Re-enable torchdynamo tests (#86297)

yanboliang committed 2 years ago

e6102881

[ao] fixing public v private for fuse_modules.py (#86028)

HDCharles committed 2 years ago

e8d3b720

[ao] fixing public v private for torch/ao/quantization (#86027)

HDCharles committed 2 years ago

d29912cc

Revert "Relandx3 "SymIntify cat and narrow" (#86289)"

pytorchmergebot committed 2 years ago

65b40807

Revert "Symintify getitem and add the required helper functions (#86207)"

pytorchmergebot committed 2 years ago

5b69b87d

Revert "Merge more symbolic meta kernels and symint changes from branch (#86334)"

pytorchmergebot committed 2 years ago

75df4b5e

Fix memory leak in _LRScheduler.step() (#85602)

KinglittleQ committed 2 years ago

b3fdb02f

Revert "Cleanup PT-D imports (#85781)"

pytorchmergebot committed 2 years ago

0e639ff4

COO intersection primitives : fusing value selection with value intersection. (#86269)

nikitaved committed 2 years ago

9b2ea41f

[autocast] Clean up registrations using new macros (#86403)

zou3519 committed 2 years ago

e125baf9

[autocast] Make it easier to register rules (#86402)

zou3519 committed 2 years ago

9b74267e

remove unused arg from `impl_func_cum_ops` (#86364)

suprajb committed 2 years ago

55f5e0de

Relandx3 "SymIntify cat and narrow" (#86289)

ezyang committed 2 years ago

a00f8489

Update distributed.rst backend collective support chart (#86406)

H-Huang committed 2 years ago

cc9183eb

[fix] sum_to_size: MathBits test - don't reuse same input tensor (#86378)

kshitij12345 committed 2 years ago

b74ca31b

Override Quantized Backend to use Fbgemm in Qlinear Packed Params Test (#86236)

salilsdesai committed 2 years ago

facbddb9

[Profiler] record gradient from nnModule (#86355)

slgong-fb committed 2 years ago

dbea07b6

Fix col2im and im2col decompositions (#86426)

lezcano committed 2 years ago

28a0b3fb

Improve interpolate() speed for channels_last images and masks (#86361)

NicolasHug committed 2 years ago

93b2d991

Fix the performance issue that the for-loop before ExternallCall could not be parallelized. (#85056)

EikanWang committed 2 years ago

70c6a988

Revert "Revert "Revert "SymIntify cat and narrow (#86191)"" (#86289)"

pytorchmergebot committed 2 years ago

2110c894

[CuDNN v8 API][Quantization]fix alignment function in quantized cuDNN V8 path (#86253)

eqy committed 2 years ago

6c604c92

Newer Older