strisunshinewentingwang/pytorch

Pull Requests Commits

[FSDP][BE] Change assert to assertEqual (#73787)

awgu committed 2 years ago

15a799e3

[Easy][FSDP] Fix warning render (#73786)

awgu committed 2 years ago

89b16f1a

[FSDP] Add grad accumulation without `no_sync()` (#73535)

awgu committed 2 years ago

4a06b8d3

[Static Runtime] Add out variant wrapper for aten::ones (#73851)

d1jang committed 2 years ago

71961d37

ci: Migrate windows conda to GHA

seemethere committed 2 years ago

2acf9c74

ci: Enable long paths for windows

seemethere committed 2 years ago

07410207

CMake option for using static MKL libraries

ashari4 committed 2 years ago

7ed73b28

[Quant][fx] Add lowering for functional conv (#73708)

andrewor14 committed 2 years ago

f3c6e8f7

[Quant][fx] Add lowering for Linear-Bn1d in QAT mode (#73509)

andrewor14 committed 2 years ago

cedce3be

[GHA] Migrate win/linux binary-smoke workflows from CircleCI

malfet committed 2 years ago

9929a9fc

HaoZeke committed 2 years ago

979a78f8

[Metal] Use MPSNNNeuronDescriptor for Mac Catalyst Target (#73718)

xta0 committed 2 years ago

1d497114

[Metal] Remove unused shader code (#73636)

xta0 committed 2 years ago

e118b1df

Revert "[GHA] Migrate win/linux binary-smoke workflows from CircleCI"

pytorchmergebot committed 2 years ago

ea698c14

Revert D34230284: [pytorch][PR] [WIP][Kineto] Manual Submodule Update

malfet committed 2 years ago

e7730eae

[GHA] Migrate win/linux binary-smoke workflows from CircleCI

malfet committed 2 years ago

486bd9f3

[WIP][Kineto] Manual Submodule Update (#73090)

aaronenyeshi committed 2 years ago

c088c8fe

Automated submodule update: FBGEMM (#73814)

facebook-github-bot committed 2 years ago

d047e475

[XNNPACK] Expose available() to xnnpack namespace (#73617)

digantdesai committed 2 years ago

4b8106f9

[Metal] Raise the minimum support version of iOS to 11.0 (#73635)

xta0 committed 2 years ago

5dbec7c0

CUDA Kernels: Use per-operator headers (1/4) (#71212)

peterbell10 committed 2 years ago

723ba4e3

CPU Kernel: Use per-operator headers (#71137)

peterbell10 committed 2 years ago

5816a95c

ns for fx: add ability for fp16 model to shadow fp32 model (#73785)

vkuzo committed 2 years ago

6f2dad24

Re-enable Windows debug libtorch

malfet committed 2 years ago

bebfdca0

[quant][fix] Fix bug for ave pooling in FX quant (#73054)

terrychenism committed 2 years ago

5167e9d5

Use higher timeout for test_tensorpipe_set_default_timeout (#73771)

pritamdamania committed 2 years ago

aca4d02d

Lock thread/block computation (#73800)

Elias Ellison committed 2 years ago

52ccbf44

[ONNX] Rewrite linspace symbolic

BowenBao committed 2 years ago

4a74285e

Revert D34554432: Back out "Revert D34524207: [pytorch][PR] remove _s_where"

ngimel committed 2 years ago

55525632

[PyTorchEdge] Add getFileFormat() so we can differentiate Zip/Pickle from Flatbuffer (#73707)

dbort committed 2 years ago

7b51629c

Newer Older