strisunshinewentingwang/pytorch

Pull Requests Commits

Simplify ScriptModule bindings. (#29432)

zdevito committed 5 years ago

4e4e29a5

switching to a simple/full executor

Krovatkin committed 5 years ago

5b702ab5

Re-enable TestNamedTensor.test_big_tensor_repr (#29407)

zou3519 committed 5 years ago

cedca377

Disabling 'contig' in quantized arithmetic test

z-a-f committed 5 years ago

b3b8f522

per-tensor quantize/dequantize benchmarking

z-a-f committed 5 years ago

5b43becf

Enable test_stress_light_rpc in rpc_test.py

mrshenli committed 5 years ago

c49b324c

Enable test_py_rref_args_user_share in rpc_test.py

mrshenli committed 5 years ago

bb90c187

Enable test_multi_py_udf_remote in rpc_test.py

mrshenli committed 5 years ago

b885eff4

Enable test_py_built_in in rpc_test.py

mrshenli committed 5 years ago

bc4457f5

Allow to create local RRef with value (#28948)

aazzolini committed 5 years ago

93b5c9d7

Add python API for get_gradients() method. (#28926)

pritamdamania committed 5 years ago

17b0ab47

qadaptive_avgpool2d benchmarking

z-a-f committed 5 years ago

9276cd44

Simple distributed optimizer (#29304)

aazzolini committed 5 years ago

b0cf43b2

F::embedding, F::embedding_bag, moved Embedding and EmbeddingBag options to embedding.h in options

anjali411 committed 5 years ago

604fc9ec

explicitly provide memory format when calling to clone() at ProcessGroupGloo.cpp

ifedan committed 5 years ago

65f3b98c

Properly shutdown RPC even in the case of `clean_shutdown=False`. (#29148)

pritamdamania committed 5 years ago

310343e9

move type inference for arange into c++ (#27629)

eellison committed 5 years ago

e01fc56e

Updating submodules

svcscm committed 5 years ago

9de0b635

Declare Dimname's kWildcard as extern instead of static (#29384)

peterbell10 committed 5 years ago

72eff002

Delete USE_CUDA macro use from data_parallel.h (#29483)

ezyang committed 5 years ago

344e7c26

Give a better error message when people accidentally use unsupported devices (#29409)

ezyang committed 5 years ago

b141754b

Move torch.cuda's atfork handler into C++ (#29101)

peterbell10 committed 5 years ago

bb119d95

Support softmax with D == 0 (#29167)

olittle committed 5 years ago

be757957

Fix missing space in lr_scheduler warning msg

adamjstewart committed 5 years ago

23483406

Disable flaky RPC tests (#29485)

mrshenli committed 5 years ago

3e5af226

Make smoke tests depend on s3 html update, to avoid race condition. (#29481)

ezyang committed 5 years ago

f6f428b6

Fix BC CI (#29533)

houseroad committed 5 years ago

46c4ae57

Revert D18286473: Use NNPACK for strided convolutions.

houseroad committed 5 years ago

466ab93e

Use handle pool to manage cuparse handles (#29426)

zasdfgbnm committed 5 years ago

2032482e

qavgpool benchmarking

z-a-f committed 5 years ago

5c9eae07

Newer Older