strisunshinewentingwang/pytorch

Pull Requests Commits

Remove TensorImpl::is_variable, deprecate Tensor::is_variable (#29653)

ezyang committed 4 years ago

65bb34d8

Only print original SourceRange on highlight

James Reed committed 4 years ago

8d23f7a3

Make sure we only run Profiling Graph Executor tests on windows (e.g. no simple, no legacy)

Krovatkin committed 4 years ago

7f4d4254

Fix tracing of autograd functions

James Reed committed 4 years ago

90ac35b7

minir edit to fix benchmark_all_test cuda error (#29829)

mingzhe09088 committed 4 years ago

747233e3

AdaptiveAvgPooling nhwc cuda update (#29700)

jjsjann123 committed 4 years ago

c5ac70a0

fix benchmark_all_test when running on gpu (#29818)

mingzhe09088 committed 4 years ago

ad95099f

add embeddingbag operator the the benchmark suite (#29784)

mingzhe09088 committed 4 years ago

b70d5712

add addmm op to the benchmark suite (#29783)

mingzhe09088 committed 4 years ago

e53b5107

Replace `make` with `cmake --build .` in the docs (#29798)

peterjc123 committed 4 years ago

dfa9c9e2

Fix typo: Caffe2_MAIN_LIB to Caffe2_MAIN_LIBS (#29746)

ezyang committed 4 years ago

01d76145

Add quantized conv3d function (#29686)

xiaomengy committed 4 years ago

bf806645

Updating submodules

svcscm committed 4 years ago

2d7d53cd

Allow rpc.remote to create RRef on self (#29634)

mrshenli committed 4 years ago

4a1fcc0b

Disable Caffe2 RCCL tests (#29792)

bddppq committed 4 years ago

9fd7db61

Update CODEOWNERS for distributed rpc framework. (#29788)

pritamdamania committed 4 years ago

ba74be0d

Enabling intra-op parallelism for fbgemm_linear_int8_weight_fp32_activation op (#29532)

jianyuh committed 4 years ago

4a27d2be

fix op benchmark OOM issue (#29794)

mingzhe09088 committed 4 years ago

f3b15727

Subscribe for record function and if android do atrace (#28708)

IvanKobzarev committed 4 years ago

aa6e992f

Use F::*FuncOptions for embedding/embeddingbag functionals (#29673)

Will Feng committed 4 years ago

a68c5249

Use NNPACK for strided convolutions. (#29595)

Ashkan Aliabadi committed 4 years ago

9ee6fa01

Linearizable Label: Class Weights, Allow Missing Label, and Average by Batch Size (#29707)

Yangxin Zhong committed 4 years ago

ed788ec7

Add error message if CUDA startup fails (#29670)

christycylee committed 4 years ago

b8dca04f

Add pytorch_jni_lite for lite interpreter. (#29621)

xcheng16 committed 4 years ago

5654eccf

use new overload mechanism for rnns (#29614)

Elias Ellison committed 4 years ago

681b610f

Simplify copy kernel with static_cast_with_inter_type (#29631)

zasdfgbnm committed 4 years ago

91bef3d1

Add more tests for torch::arange

Will Feng committed 4 years ago

65f691f2

Use default dtype for torch::tensor(floating_point_values) and torch::tensor(empty braced-init-list) when dtype is not specified (#29632)

Will Feng committed 4 years ago

2bcac59a

refactor and move createException function (#29605)

rohan-varma committed 4 years ago

3fb9bbc9

enable back 2 tests for simple exec

Krovatkin committed 4 years ago

78bd0069

Newer Older