strisunshinewentingwang/pytorch

Pull Requests Commits

[ROCm] Enable BFloat16 type for pooling ops (#34166)

rohithkrn committed 4 years ago

e907128c

ONNX Export Support for NLLLoss (#33509)

lara-hdr committed 4 years ago

8216d9ae

[pytorch][CI] add e2e mobile custom build jobs to CI (#34184)

ljk53 committed 4 years ago

e642a65b

[test all] Back out "Revert D20171428: [profiler] fix chrome tracing for profiler run with cuda"

rohan-varma committed 4 years ago

d98bd5e1

Disable MNIST test in test_xla() (#34261)

mrshenli committed 4 years ago

4a194f89

[CUDA_FUSER] Fork CUDA fuser (#33527)

jjsjann123 committed 4 years ago

2b79bab0

[JIT] fix alias assertion (#34268)

Elias Ellison committed 4 years ago

e132047f

Run RPC JIT tests with variable type hints only in Python >=3.6 (#34284)

xush6528 committed 4 years ago

e2ddf935

Add test to verify dist_autograd doesn't populate .grad field. (#33949)

pritamdamania committed 4 years ago

c62de428

Clean warning message (#34143)

VitalyFedyunin committed 4 years ago

e1c6f93f

[pytorch_ci] Don't run determination tests in py35

yns88 committed 4 years ago

1546d2af

[quant] Run weight_post_process for QAT (#33852)

supriyar committed 4 years ago

e236e159

Revert D20194092: Add support to dump unsupported ops. Add lite_interpter_load test.

mrshenli committed 4 years ago

d59e036f

Add support to dump unsupported ops. Add lite_interpter_load test. (#34072)

kimishpatel committed 4 years ago

17a5c677

[pytorch][cmake] improve build mobile with host toolchain (#34187)

ljk53 committed 4 years ago

385067ed

Make use of our S3 mirror if Yann Lecunn's website is not accessible (#34215)

ezyang committed 4 years ago

93990bab

Fix MKLDNN conv2d 5d weight handling (#34115)

Dmytro Dzhulgakov committed 4 years ago

67608cc0

[ATen] Exclude CUDA tests when running `basic` under valgrind (#34181)

malfet committed 4 years ago

9dd5d51b

Added nullptr check for pthradpool_get_threads_count (#34087)

kimishpatel committed 4 years ago

8269c4f3

Revert D20195053: [pytorch][PR] Add API for listing functions overridable by __torch_function__

mrshenli committed 4 years ago

ac6e75a1

[Dist Autograd][Better Engineering] Enhanced Error Reporting in Dist Autograd/RPC (#34179)

osalpekar committed 4 years ago

78b81dad

[torch] Fix sign-compare warning in `torch::utils::rnn:pack_sequence` (#34185)

malfet committed 4 years ago

45b8c8db

optimize UpSampleNearest 1d 2d and 3d performance on CPU (#31452)

mingfeima committed 4 years ago

39f78db7

Remove the use of macros when defining division between integers (#34104)

xuhdev committed 4 years ago

112cecc4

Cleaner implementation of bitwise operations of integeral types (#33849)

xuhdev committed 4 years ago

438f4ea0

Use templates instead of macros when defining bitwise operators. (#33835)

xuhdev committed 4 years ago

3a3fcbbc

Fix Lint (#34218)

mrshenli committed 4 years ago

78ad3dc1

[quant][graphmode] Add add_relu pattern in skip values (#32816)

jerryzh168 committed 4 years ago

6f52562e

Reduce code duplication in OperatorEntry by keying hash map on optional<DispatchKey> (#33817)

ezyang committed 4 years ago

22506ae7

Minor fix for quantizing the Ads complex model

csummersea committed 4 years ago

c688eb28

Newer Older