strisunshinewentingwang/pytorch

Pull Requests Commits

Merge branch 'master' of github.com:pytorch/pytorch into bfloat-elementwise2

zasdfgbnm committed 3 years ago

6747b205

[ROCm] update GPG key URL in circleci Dockerfile (#46256)

jeffdaily committed 3 years ago

dec61f93

[quant] Remove prehook option (#46292)

jerryzh168 committed 3 years ago

53316e8b

[ONNX] Preprocess index_put with bool inputs to masked_scatter/masked_fill (#45584)

shubhambhokare1 committed 3 years ago

9d389b1d

[quant][graphmode][fx] Move custom_module_class config to prepare/convert_custom_config_dict (#46251)

jerryzh168 committed 3 years ago

49903a5c

Update optim.rst for better understanding (#45944)

yangsenius committed 3 years ago

e7dbaa25

adding BAND/BOR/BXOR reduce ops to unsupported list for complex numbers. added tests (#46270)

bdhirsh committed 3 years ago

1f791c06

Added scalar lists APIs for addcdiv and addcmul (#45932)

Iurii Zdebskyi committed 3 years ago

8a074af9

Undefine bool and vector after including altivec.h (#46179)

Flamefire committed 3 years ago

f2e5ae4b

Remove Python version upper boundary check (#46315)

malfet committed 3 years ago

45de2ee3

Fix device guard for c10-full ops (#46091)

smessmer committed 3 years ago

69e152e6

Fix NativeFunctions.h for c10-full ops (#46090)

smessmer committed 3 years ago

4534bf57

[caffe2] Add 10s deadline for all Caffe2 hypothesis fuzz tests

malfet committed 3 years ago

84771fc6

add size_based_partition final (#46282)

scottxu0730 committed 3 years ago

62d37b9f

[jit] support tracing tensor __setitem__ with dynamic shape (#45828)

ppwwyyxx committed 3 years ago

b64cf93f

Revert D24232288: [fx] make sure args/kwargs are immutable

Mike Ruberry committed 3 years ago

38e64cf9

[JIT] Update comment in jit_log.h. (#46301)

Mikhail Zolotukhin committed 3 years ago

d790ec6d

[dispatcher] avoid autograd fixup step on non-backend keys (#46135)

Basil Hosmer committed 3 years ago

d2245512

[fx] make sure args/kwargs are immutable (#46121)

zdevito committed 3 years ago

61df99b7

[NCCL] Provide additional information about NCCL error codes. (#45950)

rohan-varma committed 3 years ago

965046c4

Only populate grad accumulator to var mapping for find_unused_parameters=True in DDP (#45942)

rohan-varma committed 3 years ago

f7398759

Parallelize the quantization conversion operators (#45536)

Radhakrishnan Venkataramani committed 3 years ago

31bcd963

Performance fix for torch.cat operator on ROCm (#46097)

ashishfarmer committed 3 years ago

d5ca53c9

[FX] Allow tracing free functions (#46268)

James Reed committed 3 years ago

09842a44

Fixed usage of std::move function (#46199)

olegfaust committed 3 years ago

ac3f23de

Use tensor's quantized properties directly in pickler (#46267)

iseeyuan committed 3 years ago

173363f3

[caffe2] Add operator schema for FP16SparseNorm (#46300)

malfet committed 3 years ago

1fcec6e7

Allow RPC framework to use rank in addition to WorkerInfo and name. (#46221)

pritamdamania committed 3 years ago

f89498f3

Reformat ivalue_inl.h and ivalue.h (#46174)

Yi Wang committed 3 years ago

e1c9aa91

[NCCL] Fix Hang in Async Error Handling due to Work logging (#46265)

osalpekar committed 3 years ago

952dc7ed

Older