strisunshinewentingwang/pytorch

Pull Requests Commits

[jit] fix traced training attribute (#47211)

zdevito committed 3 years ago

ecfa7a27

Add benchmark for per channel tensor quantization (#46017)

ajliu committed 3 years ago

27f4a78b

For torch::jit::module's attr method to moble::module (#47059)

kimishpatel committed 3 years ago

82b74bd9

[PT] optional -> c10::optional (#47144)

Hao Lu committed 3 years ago

b6685d38

[quant][graphmode][fx][fix] Linear work with float_qparam_dynamic_qconfig (#47068)

jerryzh168 committed 3 years ago

be2e3dd2

Add scalar.conj() and update backward formulas for add and sub (#46596)

anjali411 committed 3 years ago

cedeee2c

Port CPU Trace from TH to ATen (#47126)

zou3519 committed 3 years ago

86151da1

Add test for trace (#47125)

zou3519 committed 3 years ago

8054ae3e

Enable inlining into reductions (#47020)

navahgar committed 3 years ago

f58842c2

Add RAII DetectAnomalyGuard (#47164)

t-vi committed 3 years ago

b5a1be02

Remove travis-python references as well as some unnecessary dependencies (#47209)

janeyx99 committed 3 years ago

ebf36ad3

Make "Run flake8" step always succeed again (#47236)

samestep committed 3 years ago

42b6f967

Add `inputs` argument to `autograd.backward()` (#46855)

soulitzer committed 3 years ago

f5073b0c

Fix max_pool1d on discontiguous tensor (#47065)

heitorschueroff committed 3 years ago

18470f68

Revert D24335982: explicitly error out in comparison ops when the types don't match

bdhirsh committed 3 years ago

b3eb0c86

[Metal] Add pin_memory check in empty_strided (#47228)

xta0 committed 3 years ago

7f125bca

Make conversions explicit (#46835)

r-barnes committed 3 years ago

e0382065

Enhance the torch.pow testcase for the complex scalar base (#47101)

RockingJavaBean committed 3 years ago

22b3d414

annotate a few torch.nn.modules.* modules (#45772)

guilhermeleobas committed 3 years ago

9b526546

Add vulkan clamp op (#47196)

SS-JIA committed 3 years ago

71787903

add sandcastle device type test base discovery (#47119)

Rong Rong committed 3 years ago

96b23f7d

[c10] make intrusive_ptr available as a pybind holder type (#44492)

wanchaol committed 3 years ago

70d58031

[RFC] Better error message in case operator could not be run (#46885)

dhruvbird committed 3 years ago

6852cbb9

Add bfloat support for torch.randn and torch.norm (#47143)

malfet committed 3 years ago

c5ae8751

explicitly error out in comparison ops when the types don't match (#46399)

bdhirsh committed 3 years ago

60fea510

[MLF] Allow for computing prune quantile thresholds on absolute value of indicators in distributed-inference-compatible embedding LUT pruning (#46789)

Yen-Jung Chang committed 3 years ago

6e22b600

[ROCm] enable stream priorities (#47136)

jeffdaily committed 3 years ago

6906701b

Check CUDA kernel launches (fbcode/caffe2/aten/src/ATen/native/cuda/) (#47207)

Francisco Javier Ponce committed 3 years ago

c2e12333

add rocm 3.9 to nightly builds (#47121)

jeffdaily committed 3 years ago

0d6bf886

Add complex backward support for torch.exp (#47194)

anjali411 committed 3 years ago

da26858c

Newer Older