strisunshinewentingwang/pytorch

Pull Requests Commits

Skip RPC profiling tests (#38045)

rohan-varma committed 4 years ago

e3fcc6ad

[WIP][JIT] Add JIT backend registration API (#35833)

eellison committed 4 years ago

d5df055b

Add preprocessing that fuses decomposed linear into linear. (#37937)

kimishpatel committed 4 years ago

002f5ec5

Fix dummy typo in `skipIfNoFBGEMM` (#38058)

malfet committed 4 years ago

376c9a40

Fix torch.tensor dtype inference (#38030)

anjali411 committed 4 years ago

a42616f7

[TensorExpr] simplify trivial adds/subs/muls even in Float (#37960)

nickgg committed 4 years ago

f2f80277

Back out "Revert D18927220: if_constexpr for C++14" (#37792)

smessmer committed 4 years ago

379e717a

stop creating integer type Tensors that require gradients (#37789)

albanD committed 4 years ago

5e83a13e

Make profiler thread local (#36291)

Ilia Cherniavskii committed 4 years ago

facc5e0c

Propagate TLS callbacks with ThreadLocalState (#37745)

Ilia Cherniavskii committed 4 years ago

2ef40105

Move RecordFunction into ATen (#37548)

Ilia Cherniavskii committed 4 years ago

2d708cef

Make RecordFunction callbacks thread local and modernize interface (#37491)

Ilia Cherniavskii committed 4 years ago

c24c5f96

Move resize / zero logic for _thnn_conv_depthwise2d from codegen to native code. (#37957)

gchanan committed 4 years ago

dc251908

Move _thnn_conv2d resize and zero code from codegen to native code. (#37956)

gchanan committed 4 years ago

ed4e7cec

Fixed gradcheck for complex (#37836)

anjali411 committed 4 years ago

99349393

Remove unpacked int8 blob after constructing the packed blob to save memory (#37973)

csummersea committed 4 years ago

8a8b7a16

[Tensorpipe Agent] Implementing getMetrics with currently available metrics (#37980)

osalpekar committed 4 years ago

f0f58736

[Tensorpipe Agent] Network Data Profiling (#37852)

osalpekar committed 4 years ago

5d21a9cf

[Tensorpipe Agent] Implement Global Interpreter Lock Wait Time Metric (#37851)

osalpekar committed 4 years ago

25359f73

[Tensorpipe Agent] Base Structs for Tracking RPC Metrics (#37850)

osalpekar committed 4 years ago

b452fef5

[TensorExpr] Allow to enable/disable fallback mechanism thru an envvar PYTORCH_TENSOREXPR_FALLBACK. (#37971)

Mikhail Zolotukhin committed 4 years ago

a44824c9

[TensorExpr] Move controlling knob out of the TE fuser pass. (#37970)

Mikhail Zolotukhin committed 4 years ago

067f08c1

Remove overly strict assertion for type demotion of scalars. (#38001)

resistor committed 4 years ago

3066d3ac

[quant] Release qnnpack original weights for conv/linear (#37595)

supriyar committed 4 years ago

7bf9d983

Make speed_benchmark_torch report latency in us (#37953)

kimishpatel committed 4 years ago

dd64d26d

Message Delay fix for test_check_failed_messages (#37978)

osalpekar committed 4 years ago

85fccba2

Update miniconda repository, be specific about cudatoolkit (#37186)

seemethere committed 4 years ago

305444a0

[BE] Add @skipIfNoFBGEMM decorator (Reland) (#37894)

malfet committed 4 years ago

2b41b9bc

[CUDA_FUSER] Expand operation support for cuda fuser (#37849)

jjsjann123 committed 4 years ago

1667aa64

[TensorPipe] Update submodule (#38013)

lw committed 4 years ago

ffed9dca

Newer Older