strisunshinewentingwang/pytorch

Pull Requests Commits

Name inference rules for relu/relu_/threshold/threshold_ (#25569)

zou3519 committed 5 years ago

0ebbcd95

Fix named tensor printing (#25564)

zou3519 committed 5 years ago

9ea6238b

Add the dynamic quantized LSTM module (#25157)

jianyuh committed 5 years ago

0483d537

Fuse to individual operators to GatherFuse8BitRowwiseQuantFloatMulLengthElim (#25519)

Yinghai Lu committed 5 years ago

4edf77b6

change shape for some ops to reduce variance

hl475 committed 5 years ago

cd4a7cda

Fix binary op name inference to happen before shape checks (#25563)

zou3519 committed 5 years ago

67d64ea9

Name inference rule for torch.cat (#25568)

zou3519 committed 5 years ago

9922e094

Eliminate magic numbers in BatchLinearAlgebra.cu (#25524)

vishwakftw committed 5 years ago

49baeb9d

Don't save `self` in `index` backward (#25594)

zou3519 committed 5 years ago

8edf149f

Name inference for masked_fill_ / masked_fill

zou3519 committed 5 years ago

a6ba4f64

Name inference rule for masked select (#25566)

zou3519 committed 5 years ago

2aef6066

Enable torch.cholesky for batches > 262140 (#24438)

vishwakftw committed 5 years ago

d1e079e2

Get rid of _th_reciprocal_. (#25507)

gchanan committed 5 years ago

0621e2ce

Enable broadcasting of batch dimensions RHS and LHS tensors for lu_solve (#24333)

vishwakftw committed 5 years ago

1e4832ff

Updating submodules

svcscm committed 5 years ago

914a6051

Documentation for cdist (#25221)

ifedan committed 5 years ago

896cd1c5

Remove index calculation in quantized max_pool2d (#25526)

James Reed committed 5 years ago

9cb9f159

Make scatter/gather arguments optional (#25575)

pietern committed 5 years ago

500e72aa

Error phrasing in torch.distributed helper functions

pietern committed 5 years ago

493f7bd8

Name inference rule for mean, std, var, std_mean, var_mean (#25431)

zou3519 committed 5 years ago

938e7402

Added invert bitwise operation to JIT (#22324)

Chillee committed 5 years ago

f3f83ccb

Prepare to add some Dimname/DimnameList overloads (#25405)

zou3519 committed 5 years ago

5c4cc1e8

Migrate multinomial from the TH to ATen (CPU) (#25274)

ifedan committed 5 years ago

c89301a6

Implement indexing methods for sparse tensors (#24937)

pearu committed 5 years ago

f793a7c5

Update index.rst (#24245)

brianjo committed 5 years ago

832c72a2

Create helpers for implementing unary ops whose CUDA implementation is ATen. (#24879)

xuhdev committed 5 years ago

b46bc79f

Delete torch/csrc/nn/type_checks, which aren't used anymore.

gchanan committed 5 years ago

4864403f

Add copy logic for LibTorch to avoid issues on Windows (#25556)

peterjc123 committed 5 years ago

09ef107e

Updating submodules

svcscm committed 5 years ago

ba9f1344

add options to flush cache in SLS benchmarks (#25530)

jspark1105 committed 5 years ago

8199bb3d

Newer Older