strisunshinewentingwang/pytorch

Pull Requests Commits

IRParser: optionally create name->value map of the parsed IR. (#19551)

Mikhail Zolotukhin committed 5 years ago

8abab61d

Profiling : Adding Profile Op to provide storage for profiling lambdas

Krovatkin committed 5 years ago

43d0b78c

Step 5: remove _unique_dim in favor of unique_dim (#18654)

zasdfgbnm committed 5 years ago

7f053b27

Add back option to not adjust output batch size (#19442)

Yinghai Lu committed 5 years ago

767d184b

Add debug logic to c2_ref_test and its helpers (#19359)

Michael Antonov committed 5 years ago

7655b857

fix variable shadowing issus (#19567)

Dehua Cheng committed 5 years ago

a09240b0

Add manual_seed in script (#19510)

Elias Ellison committed 5 years ago

19f73180

Automatic update of fbcode/onnx to 83dd62659fc07d5b7fa93b5d1c1879f93509c7db (#19454)

houseroad committed 5 years ago

e714429b

Get rid of unnecessary matches_jit_signature: True specifications. (#19549)

gchanan committed 5 years ago

5d5d67fa

Rename potri to cholesky_inverse (#19498)

vishwakftw committed 5 years ago

c30224ad

Add assertion to make sure init op is always fp16 compatible in fp16 training

chocjy committed 5 years ago

deadf3ba

Generate only one Type class per backend (#19295)

Roy Li committed 5 years ago

689dd800

Make complex its own backend (#19275)

Roy Li committed 5 years ago

189f3060

Add ScalarType argument to Type::options() (#19270)

Roy Li committed 5 years ago

ab78449e

Generate cases for all ScalarTypes in Type functions that call to TH (#19230)

Roy Li committed 5 years ago

a044ba1a

Fix clang-format. (#19550)

Mikhail Zolotukhin committed 5 years ago

868933a4

Fix some typos in jit README

mrshenli committed 5 years ago

1bd5f2c1

Match JIT signature with triu_indices / tril_indices. (#19484)

gchanan committed 5 years ago

5afc2747

Make one_hot non-differentiable. (#19524)

gchanan committed 5 years ago

9eb48e1b

Remove 'BoolTensor', 'IndexTensor' from frontend specifications. (#19523)

gchanan committed 5 years ago

67330374

Have _embedding_bag_dense_backward match JIT signature. (#19522)

gchanan committed 5 years ago

39446015

Have embedding_dense_backward match JIT signature. (#19521)

gchanan committed 5 years ago

e3523979

Update mkldnn-bridge to fix crash issue in DNNLOWP dequantize op (#19159)

gujinghui committed 5 years ago

638ffac3

Hook up non_differentiability in derivatives.yaml when no autograd function is generated. (#19520)

gchanan committed 5 years ago

83373e77

Move non_differentiable_arg_names from autograd functions to differentiability_info. (#19519)

gchanan committed 5 years ago

8868a4f2

Move cuFFT plan cache note outside Best Practices (#19538)

ssnl committed 5 years ago

6d307db5

Revert D14689639: [pytorch] Allow passing lists as trace inputs.

suo committed 5 years ago

26f1c6d4

Improve optimizations for DNNLOWP support on MKL-DNN (#18843)

gujinghui committed 5 years ago

c96c91da

Make Observer class as template Quant class for QuantConfig (#19418)

nishantpdce committed 5 years ago

fe87327c

Support compilation on gcc-7.4.0 (#19470)

Sam Leeman-Munk committed 5 years ago

9f4f7e16

Newer Older