strisunshinewentingwang/pytorch

Pull Requests Commits

Dynamic quantization support for LSTMCell, RNNCell and GRUCell (#37159)

raghuramank100 committed 4 years ago

48db06e3

Change AccumulateGrad to yield `.grad`s that match weights' memory layout (#34904)

definitelynotmcarilli committed 4 years ago

2beb9690

Add smoke test to Windows CI (#39941)

guyang3532 committed 4 years ago

c0650495

Eliminate TensorIterator::add_output with explicit dtype. (#39800)

ezyang committed 4 years ago

d71804a5

[DataLoader] add repr for WorkerInfo (#39975)

ssnl committed 4 years ago

23db54ac

[quant][graphmode] Pass debug option into insert_quant_dequant pass (#39915)

jerryzh168 committed 4 years ago

ee5ad6ce

Clarifies compare_with_numpy behavior (#40064)

Mike Ruberry committed 4 years ago

ebd86915

Revert D21986243: TORCH_FN

Mike Ruberry committed 4 years ago

8939849f

TORCH_FN (#39823)

smessmer committed 4 years ago

12cb80b5

[quant][graphmode] Use quantizedbatch_norm in graph mode (#39911)

jerryzh168 committed 4 years ago

144e8dc5

Refactor LSTM tests (#38851)

raghuramank100 committed 4 years ago

655f1ea1

[pytorch] consolidate android gradle build scripts (#39999)

ljk53 committed 4 years ago

bcb44796

Back out "[pytorch][PR] Remove `THTensor_(fill)` & `THTensor_(zero)`"

ngimel committed 4 years ago

9204d76b

[pytorch] simplify android circleci definition data model

ljk53 committed 4 years ago

f0b40cac

Revert D22050656: [Yet Another Reland] Include AT_PARALLEL_OPENMP/AT_PARALLEL_NATIVE/AT_PARALLEL_NATIVE_TBB to ATen/Config.h

pbelevich committed 4 years ago

18fe9d26

[quant] add quantized::batch_norm (#39910)

jerryzh168 committed 4 years ago

1a388da1

DNNL: fix F.max_pool2d and F.avg_pool2 issue when stride=None (#39221)

XiaobingSuper committed 4 years ago

5d4a6628

Fix TensorPipeAgent shutdown to ensure it drains all outstanding work. (#40060)

pritamdamania committed 4 years ago

399dd84c

[Yet Another Reland] Include AT_PARALLEL_OPENMP/AT_PARALLEL_NATIVE/AT_PARALLEL_NATIVE_TBB to ATen/Config.h (#40045)

pbelevich committed 4 years ago

d4faf14c

Check if generator has next normal sample cache methods in normal_distribution (#39816)

pbelevich committed 4 years ago

f13be5fd

[distribtued.nn] Implement TorchScript-compatible RemoteModule API (#37139)

xush6528 committed 4 years ago

00651b8c

[quant][graphmode] Support prim:TupleUnpack and prim::TupleConstruct (#39895)

jerryzh168 committed 4 years ago

f37b8e73

Overload complex math functions on both :: and std:: (#39829)

zasdfgbnm committed 4 years ago

eb358f49

[android] Remove android fbjni subproject (#39691)

IvanKobzarev committed 4 years ago

84d8a42f

[torch.distributed.rpc] Add WorkerInfo python repr magic method (#40004)

xush6528 committed 4 years ago

d602950c

[TensorPipe] Use registry for transports and channels (#39957)

lw committed 4 years ago

ecfe0c9a

[bernoulli_kernel] Replace CPU_tensor_apply functions with cpu_serial_kernel (#39711)

RockingJavaBean committed 4 years ago

51e341df

[futures] Reland: Add torch.futures.collect_all()/wait_all() python api. (#39964)

jjlilley committed 4 years ago

0c254285

[resubmit] [pytorch][PR] Fix for num_threads==1 in OpenMP "parallel for" (#39533)

Ilia Cherniavskii committed 4 years ago

cc3fc786

topk tensor k support (#39407)

shz117 committed 4 years ago

f6b0fbe2

Newer Older