strisunshinewentingwang/pytorch

Pull Requests Commits

Updating submodules

svcscm committed 4 years ago

0bfe12d0

batch size 0 tests for Quantize/Dequantize DNNLOWP ops (#26873)

jspark1105 committed 4 years ago

d3cab657

batch size 0 support in FC DNNLOWP operators (#26872)

jspark1105 committed 4 years ago

78b0c58a

batch size 0 tests for element-wise DNNLOWP ops (#26870)

jspark1105 committed 4 years ago

41c1cc2f

batch size 0 support in Conv DNNLOWP ops (#26871)

jspark1105 committed 4 years ago

1aaf4810

Add bitwise distributed reduction ops (#26824)

pietern committed 4 years ago

2991bfdb

Add some missing constructors to IValue.

ezyang committed 4 years ago

dec0b6b7

Make resize_as_ generic, so XLA works. (#26809)

ezyang committed 4 years ago

60b57d96

batch size 0 support in ChannelShuffle DNNLOWP op (#26858)

jspark1105 committed 4 years ago

8fb756d3

Add more inplace arguments to quantization top level API (#26782)

Dmytro Dzhulgakov committed 4 years ago

0a8a779a

Enable batch_size = 0 support in DNNLOWP Concat operator (#26849)

jspark1105 committed 4 years ago

5231699d

remove backward functions from jit-op-registry for mobile build (#26851)

ljk53 committed 4 years ago

60372dc7

add mobile friendly at:parallel_for backend

ljk53 committed 4 years ago

ed260748

Improvements to GuardElimination and InsertBailouts

Krovatkin committed 4 years ago

14d7d571

Updating submodules

svcscm committed 4 years ago

20ed6ba0

Remove unnecessary functions and cleanup code in quantization.cpp.

Mikhail Zolotukhin committed 4 years ago

058ba0e7

Fix building with PARALLEL_BACKEND=NATIVE_TBB (#26742)

Arnaud Bonnet committed 4 years ago

8f359a48

Remove three unused declaration. (#26699)

xuhdev committed 4 years ago

c25c507f

Try to disable annoying hypothesis warnings again (#26853)

James Reed committed 4 years ago

f37aa2de

Re-write of tensor-scalar quantized add

James Reed committed 4 years ago

20ebd13f

Fix broken failure messages for OverloadedMethodValue

James Reed committed 4 years ago

1d55616a

Throw if someone tries to torch.save() quantized modules (#26828)

James Reed committed 4 years ago

df16fb9c

Remove convert_to_ssa argument from runCleanupPasses - it is only used in one place.

Mikhail Zolotukhin committed 4 years ago

d842435c

Use optimized_graph in graph_executor.

Mikhail Zolotukhin committed 4 years ago

9df887df

QEngine::QNNPACK enabled, module.eval()

IvanKobzarev committed 4 years ago

ed82a28c

Migrate multinomial from the TH to Aten (CUDA) (#26481)

ifedan committed 4 years ago

2eb59232

enable double backward for non-cudnn LSTM and GRU (#26660)

ngimel committed 4 years ago

90ffab6e

Move the CUDA implementation of log to ATen. (#26494)

xuhdev committed 4 years ago

91549ef6

Bytecode export flow (#25187)

Martin Yuan committed 4 years ago

7fc06ea5

Highlighting in the doc that square root comes before adding epsilon

vincentqb committed 4 years ago

660d9e24

Newer Older