strisunshinewentingwang/pytorch

Pull Requests Commits

[ONNX] Update slice symbolic function (#42935)

KsenijaS committed 4 years ago

820c4b05

Preserve python backtrace in autograd engine errors. (#43684)

pritamdamania committed 4 years ago

f1624b82

[reland][quant][graphmode][fx] Add support for weight prepack folding (#43728) (#43902)

jerryzh168 committed 4 years ago

825c109e

Update torch.range warning message regarding the removal version number (#43569)

xuhdev committed 4 years ago

6da26cf0

[TensorExpr] Check statements in test_kernel.cpp (#43911)

asuhan committed 4 years ago

85d91a32

Revert D23335106: [quant][graphmode][fix] Fix insert quant dequant for observers without qparams

Ailing Zhang committed 4 years ago

f229d2c0

simplify profile text output by displaying only top-level ops statistics (#42262)

YazhiGao committed 4 years ago

69080e9e

Update determinism documentation (#41692)

kurtamohler committed 4 years ago

d7ee84c9

Remained changes of #43578 (#43921)

xuhdev committed 4 years ago

69fbc705

[caffe2] Extend dedup SparseAdagrad fusion with stochastic rounding FP16 (#43124)

jianyuh committed 4 years ago

3c2f6d2e

Fix exception chaining in `torch/` (#43836)

akihironitta committed 4 years ago

f17d7a55

Move type annotations for remaining torch.utils stub files inline (#43406)

rgommers committed 4 years ago

da32bf4c

[quant][graphmode][fix] Fix insert quant dequant for observers without qparams (#43606)

jerryzh168 committed 4 years ago

60220975

[reland][quant][graphmode][fx] Add top level APIs (#43581) (#43901)

jerryzh168 committed 4 years ago

7db7da71

[TensorExpr] Make KernelSumMultipleAxes much faster (#43905)

asuhan committed 4 years ago

deb5fde5

[ONNX] Floordiv (#43022)

KsenijaS committed 4 years ago

ee53a335

Avoid resizing in MinMaxObserver (#43789)

Qi Zhou committed 4 years ago

f73ba889

[JIT] Remove loop peeling from the profiling executor pipeline. (#43847)

Mikhail Zolotukhin committed 4 years ago

98b846cd

[JIT] Specialize autograd zero: actually remove the original graph after we created its versioned copy. (#43900)

Mikhail Zolotukhin committed 4 years ago

d69d6030

[JIT] Specialize autograd zero: fix the guarding condition. (#43846)

Mikhail Zolotukhin committed 4 years ago

f150f924

Fix ImportError in the OSS land. (#43912)

venkatacrc committed 4 years ago

9b820fe9

log message at per-test level for`perfpipe_pytorch_test_times` (#43752)

Danqi Huang committed 4 years ago

71373276

Move torch/autograd/grad_mode.pyi stubs inline (#43415)

rgommers committed 4 years ago

4c19a1e3

Enable gcc coverage in OSS (#43883)

yujunzhao@devvm1621.atn0.facebook.com committed 4 years ago

e941a462

Move `fbcode` related coverage code to `fb/` folder and add `TARGETS` (#43800)

yujunzhao@devvm1621.atn0.facebook.com committed 4 years ago

da0e93a8

Implementing NumPy-like function torch.heaviside() (#42523)

Kiyosora committed 4 years ago

3682df77

Let linspace support bfloat16 and complex dtypes (#43578)

xuhdev committed 4 years ago

7680d87a

Skip target determination for codecov test (#43899)

malfet committed 4 years ago

3278beff

[pytorch][bot] update mobile op deps (#43871)

ljk53 committed 4 years ago

ffca81e3

Join-based API to support DDP uneven inputs (#42577)

rohan-varma committed 4 years ago

4e4626a2

Newer Older