strisunshinewentingwang/pytorch

Pull Requests Commits

[quant][graph][fix] Set type for GetAttr nodes in remapTypes (#46250)

supriyar committed 3 years ago

95ccf34f

Use storage.cpu() for moving storage to CPU in serialization. (#46028)

t-vi committed 3 years ago

7b7f2519

.circleci: Fix android publish snapshot job (#46266)

seemethere committed 3 years ago

fc846db6

[quant][refactor] Alphabetize the entries in the quantized import (#46218)

z-a-f committed 3 years ago

5604997b

Support pytest for distribution testing (#45648)

neerajprad committed 3 years ago

faa9c22a

trying to make pow work for tensor raised to the power of a scalar (#46185)

janeyx99 committed 3 years ago

ad376f1a

Add torch._foreach_maximum(TensorList, TensorList) & torch._foreach_minimum(TensorList, TensorList) APIs (#45692)

Iurii Zdebskyi committed 3 years ago

1a57b390

Define the record_stream method in native_functions.yaml (#44301)

chengjunlu committed 3 years ago

5741de88

Refactor dispatcher and native to use Signature structure. (#45990)

ezyang committed 3 years ago

d705083c

Remove unnecessary byte-for-byte compatibility code that is not needed. (#45975)

ezyang committed 3 years ago

f0860326

Rename legacy_dispatcher to native. (#45974)

ezyang committed 3 years ago

8d5c899b

Reorder dispatcher/legacy_dispatcher types (#45973)

ezyang committed 3 years ago

527a8bee

Add NativeFunctionGroup (#45918)

ezyang committed 3 years ago

944eb0e3

Rewrite implementation of faithful cpp signatures (#45890)

ezyang committed 3 years ago

9079aea1

fix #45552 - adding add_done_callback(fn) to torch.futures.Future (#45675)

bdhirsh committed 3 years ago

a3caa719

Workaround for bug in DistributedDataParallel (#46186)

Flamefire committed 3 years ago

282f4ab9

[iOS][GPU] Add Metal/MPSCNN support on iOS (#46112)

xta0 committed 3 years ago

a277c097

[quant][fx][graphmode][api] Change API for custom module (#45920)

jerryzh168 committed 3 years ago

7f6a1b2b

Revert D24165889: Update native_functions.yaml to add DefaultBackend.

Ailing Zhang committed 3 years ago

e6d30c89

Update native_functions.yaml to add DefaultBackend. (#45938)

Ailing Zhang committed 3 years ago

1f9ddf64

Use const-references in nodes_to_rewrite range loop

malfet committed 3 years ago

ba1e0a88

Fix JIT test config (#46230)

malfet committed 3 years ago

4ad47156

Fix incorrect CUDA `torch.nn.Embedding` result when max_norm is not None and indices are not sorted (#45248)

kurtamohler committed 3 years ago

66505b64

[fx] use a linked list for nodes (#45708)

zdevito committed 3 years ago

88dcb95e

Adding information how to control randomness with DataLoader (#45749)

VitalyFedyunin committed 3 years ago

31ee5d8d

[pytorch][PR][Gradient Compression] Reduce the peak memory of fp16 compression provided by ddp comm hook (#46078)

Yi Wang committed 3 years ago

ee3d3e6d

[pt][quant] Support either min or max in qclamp (#45937)

dskhudia committed 3 years ago

87a4baf6

Implement ravel (#46098)

ejguan committed 3 years ago

bed3b405

fix test_serialization not working with Windows. (#46120)

Rong Rong committed 3 years ago

b98e3594

[NNC] Fix two more bugs in Cuda Half support (#46129)

nickgg committed 3 years ago

f3db6877

Newer Older