strisunshinewentingwang/pytorch

Pull Requests Commits

[Selective build] Enable selective build in VariablType

iseeyuan committed 4 years ago

85128113

Assert that kernels are called with the right signature (#38361)

smessmer committed 4 years ago

55cdd31b

Get rid of global constructors in cuda codegen (#40183)

ionsphere committed 4 years ago

d14d47b9

[android] ANDROID_STL=c++_shared (#39588)

IvanKobzarev committed 4 years ago

0891764e

[android] Add libtorch headers to pytorch_android aar (#39507)

IvanKobzarev committed 4 years ago

d3b786af

.circleci: Add docker builds based on rev-parse (#40194)

seemethere committed 4 years ago

83d7718c

[test] split remaining quantization tests out of test_jit (#40144)

wanchaol committed 4 years ago

442ec1dd

Revert D22108899: Ensure NCCL_BLOCKING_WAIT=1 works for dist.barrier()

Ilia Cherniavskii committed 4 years ago

30648985

[android][cmake] Remove NO_EXPORT for libtorch mobile build (#39584)

IvanKobzarev committed 4 years ago

74a2cb87

Fix typos in RPC Docs (#40219)

gemfield committed 4 years ago

034eddca

preserve output tensor's stride in TI's fast setup (#38895)

lixinyu committed 4 years ago

645d6c01

.circleci: Move docker_build workflow to codegen (#40189)

seemethere committed 4 years ago

6a42d85f

[quant][api] Expose graph mode quantization API in `torch.quantization` (#40198)

jerryzh168 committed 4 years ago

aa84ec53

[codemod][custom_rule] Migrate some scripts to use named outputs for custom_rule

irenewchen committed 4 years ago

fef253e7

graph mode: move hardsigmoid op to `single_input_general_value` category (#40055)

vkuzo committed 4 years ago

fcc9a1e6

graph mode: util for fusion of functions which require observation (#39413)

vkuzo committed 4 years ago

37362fff

quant layer/group/instance norm: make weights and biases optional (#39203)

vkuzo committed 4 years ago

4ad8ebe7

[quant][graphmode] Add support for detach (#40197)

jerryzh168 committed 4 years ago

d4e4f131

Move the check on orig_weight sizes. (#40200)

kimishpatel committed 4 years ago

5f309505

[JIT x RPC] Consolidate RRef type class and RRef impl class (#35694)

xush6528 committed 4 years ago

f3f30d43

[TensorPipe] Add options for agent, including backend killswitches (#40162)

lw committed 4 years ago

7c9e78fd

Ensure NCCL_BLOCKING_WAIT=1 works for dist.barrier() (#40207)

pritamdamania committed 4 years ago

d1a0e880

Reduce number of Window test configurations (#38482)

malfet committed 4 years ago

4553b0b5

[quant][graphmode] Clean up and add more logging (#40196)

jerryzh168 committed 4 years ago

fd7e09a5

Move _dummy_type to _utils.py (#40177)

malfet committed 4 years ago

76fbfba6

Remove thrust::complex from sqrt (#39901)

zasdfgbnm committed 4 years ago

efd9fc74

Add aarch64 specific quantize_tensor using arm intrinsics. (#40113)

kimishpatel committed 4 years ago

edd3fbc6

Export box_cox operator in caffe2

Zhuobo Feng committed 4 years ago

fb02007e

[quant][graphmode] Add warning for prim::Loop (#40195)

jerryzh168 committed 4 years ago

18000327

Add optimization blacklist as second arg to optimizeForMobile method. (#37462)

xcheng16 committed 4 years ago

0b3755b1

Newer Older