strisunshinewentingwang/pytorch

Pull Requests Commits

Update PULL_REQUEST_TEMPLATE.md (#70105)

albanD committed 2 years ago

54406314

Remove backward ops for miopen convolution (#69987)

jbschlosser committed 2 years ago

b1d5948b

dbr quant: extend qconfig_dict support to functionals, part 2 (#69766)

vkuzo committed 2 years ago

f045618d

dbr quant: extend qconfig_dict support to functions, part 1 (#69758)

vkuzo committed 2 years ago

a4173fc8

dbr quant: make fqn during prepare op hook required (#69726)

vkuzo committed 2 years ago

c186773d

fx quant: move _parent_name to common utils (#69720)

vkuzo committed 2 years ago

b999f875

dbr quant: initial support of qconfig_dict for modules (#69719)

vkuzo committed 2 years ago

4f450f44

fx quant: refactor qconfig_dict utils to separate file (#69636)

vkuzo committed 2 years ago

0f1ceb34

[tensorexpr] Add memory planning to reuse intermediate buffers (#66452)

huiguoo committed 2 years ago

7abb7667

[tensorexpr] Remove the optional argument in LoopNest::prepareForCodeGen (#67144)

huiguoo committed 2 years ago

ac92f7cc

[tensorexpr] Move the allocation of intermediate buffers from TEK to CodeGen (#67143)

huiguoo committed 2 years ago

bbfd7b75

[tensorexpr] Add BufMap instruction to reuse the memory of dest buf for src buf (#66451)

huiguoo committed 2 years ago

6075ec15

[tensorexpr] Add a stmt recorder to obtain stmt PCs (#66450)

huiguoo committed 2 years ago

c7e09515

[quant][graphmode] Rename backend_config_dict folder to backend (#69882)

jerryzh168 committed 2 years ago

043098ef

[DataPipe] Unifying API - removing options to have fn_args and fn_kwargs from MapDataPipes (#69561)

NivekT committed 2 years ago

3d51c880

[DataPipe] Unifying API - removing options to have fn_args and fn_kwargs from IterDataPipes (#69560)

NivekT committed 2 years ago

b89c283c

OpInfos for torch.{flatten, column_stack} (#69237)

anjali411 committed 2 years ago

4a6a5d16

[quant][be] Cleanup test cases for eager mode workflow (#69880)

jerryzh168 committed 2 years ago

ef6f776e

[shard] remove set device for nccl (#69946)

wanchaol committed 2 years ago

92320dfe

[reland][quant][fx][graphmode] Add support for conv add pattern in backend_config_dict (#70007)

jerryzh168 committed 2 years ago

98136295

.github: Volume mount netrc to home directory (#70057)

seemethere committed 2 years ago

62809dc0

[reland][quant][graphmode][fx] Enable fuse handler for sequence of 3 ops (#70006)

jerryzh168 committed 2 years ago

a73c6a45

Fix lint/mypy violations (#70059)

malfet committed 2 years ago

fa582045

extract out c10 targets to the c10 package (#69992)

Michael Dagitses committed 2 years ago

02c63c30

[jit][edge] Remove usage of shared_ptr<mobile::Code>. (#68037)

zhxchen17 committed 2 years ago

d459e795

[jit] Split ClassType into a separate header. (#68036)

zhxchen17 committed 2 years ago

39f65fee

Sparse CSR CUDA: Add block sparse support for torch.triangular_solve (#68709)

IvanYashchuk committed 2 years ago

243e135e

update `SequentialLR` signature (#69817)

rohitgr7 committed 2 years ago

5f3f327a

Revert D33136054: Remove backward ops for miopen convolution

jbschlosser committed 2 years ago

15b9e5f8

Provide functionality to write custom ShardedTensor ops. (#69874)

pritamdamania committed 2 years ago

b199e3c8

Newer Older