strisunshinewentingwang/pytorch

Pull Requests Commits

migrate `gather` to ATen (CUDA) (#37659)

nikitaved committed 4 years ago

5a19fe74

use version number instead of 'master' in html header title (#38149)

glaringlee committed 4 years ago

52e9953f

Use `jit_core_sources` from build_varliables.bzl (#38526)

malfet committed 4 years ago

4b52e525

Add tan_cuda for complex dtypes (#38400)

anjali411 committed 4 years ago

242af6c0

Adding support for manifold files in DBReader (#37727)

mehtanirav committed 4 years ago

acacad25

Issue 37819: Added check for kHIP in ATen/native/Copy.cpp (#38003)

mergian committed 4 years ago

bae895ce

Add message to static_assert (#38519)

xwang233 committed 4 years ago

bf2bbd96

Revert "Vectorize non-persistent Softmax kernels (#36485)" (#38534)

ngimel committed 4 years ago

c0bc1827

[TensorExpr] Fix bug when splitting inner reduce axis with tail (#38420)

nickgg committed 4 years ago

8bf31245

Updating submodules

svcscm committed 4 years ago

0d51728d

Remove some unnecessary cast for complex numbers. (#38422)

xuhdev committed 4 years ago

3cb2778d

Support operations on c10::complex and integer scalars (#38418)

xuhdev committed 4 years ago

000fea37

Update NNPI backend to 0.5.2.5. (#4464)

omromano committed 4 years ago

ac613371

[quant][graphmode][refactor] Factor out getFixedQParamOpFusionInfo (#38359)

jerryzh168 committed 4 years ago

ec9b2f9a

[JIT] Fix `@staticmethod` access from `self` on modules (#37702)

voznesenskym committed 4 years ago

960f4b51

[c2] fix compute_norm test (#38529)

Wakeupbuddy committed 4 years ago

3d0532f3

Add sccache support for hcc and hip-clang in ROCm (#38451)

mpruthvikumar committed 4 years ago

8df14c57

Back out "[c2] register cuda op for LpNorm (fallback)"

Wakeupbuddy committed 4 years ago

fac9f365

[quant][graphmode][refactor] Factor out getInputTensorQParamOpFusionInfo (#38358)

jerryzh168 committed 4 years ago

ee525019

Enforce const on PyRRef functions (#38415)

Shen Li committed 4 years ago

155a287a

[quant] Support empty batch input for quantized ops (#38508)

supriyar committed 4 years ago

25177e27

Revert D21585458: [pytorch][PR] [RELAND] .circleci: Improve docker image build workflow

Ailing Zhang committed 4 years ago

bc49d938

[c2] [easy] stop gradient when diagnose (#38518)

Wakeupbuddy committed 4 years ago

0e0b9496

[RELAND] .circleci: Improve docker image build workflow (#38484)

seemethere committed 4 years ago

8cdc4807

[c2] register cuda op for LpNorm (fallback) (#38517)

Wakeupbuddy committed 4 years ago

bbfd0ef2

[quant][graphmode] Support ops with fixed quantization parameters (#38278)

jerryzh168 committed 4 years ago

504637a1

[quant] Support for functional quantized::conv1d (#38449)

supriyar committed 4 years ago

de7025fb

[quant][graphmode] Add support for quantized conv1d + relu fusion (#38441)

supriyar committed 4 years ago

8e732514

[quant] Fusion support for conv1d + ReLU (#38438)

supriyar committed 4 years ago

f4605ae5

Add C++ Landing Page (#38450)

jlin27 committed 4 years ago

8b6bf2a4

Newer Older