strisunshinewentingwang/pytorch

Pull Requests Commits

Fix cuda and cudnn libraries search process on Windows (#20205)

peterjc123 committed 5 years ago

d6f62b70

Remove explicit checks for parallelism from TH (#20002)

Ilia Cherniavskii committed 5 years ago

d24c0aa8

Port TH library to ATen/Parallel (#19105)

Ilia Cherniavskii committed 5 years ago

0ebe252c

Namespace isolation for classes (#19903)

suo committed 5 years ago

26dd65ea

fix parsing bugs (#20246)

zdevito committed 5 years ago

e41aa0ed

Extract repeated scripts into files (#19674)

glenjamin committed 5 years ago

21a3895c

add decay parameter in ref_adagrad (#15329)

jspark1105 committed 5 years ago

42e9a619

add str builtin support (#20188)

zdevito committed 5 years ago

f0b5ad89

Autograd profile recording in c10 custom ops (#20175)

James Reed committed 5 years ago

9fcf5854

smoke test for add_graph (#20007)

lanpa committed 5 years ago

fb9d9fbd

tweak scripts/build_android.sh for ABI and header install (#20152)

t-vi committed 5 years ago

3ac4d928

Standard gamma's export

houseroad committed 5 years ago

faf2c3ac

Automatic update of fbcode/onnx to 5bde6371620b76302864bce90f521d72eda95d0e (#20232)

houseroad committed 5 years ago

48e649b8

Support operator overloading for UDT (#20033)

Elias Ellison committed 5 years ago

6606ac5d

Remove brew libomp from binary mac machines (#20228)

pjh5 committed 5 years ago

9c62280e

Fix in benchmark_test_generator (#20237)

Ilia Cherniavskii committed 5 years ago

eecf52b4

Remove flake8 E303 (too many blank lines) (#20225)

suo committed 5 years ago

8a375189

Fix test_namedtuple_return (#20212)

zasdfgbnm committed 5 years ago

ba3e6de4

Use ignore=dirty in submodules. (#20135)

zdevito committed 5 years ago

785583a4

PyTorch Profiler Shape aggregation support (#20035)

salexspb committed 5 years ago

864cfbc2

support onnx export rnn with batch_first=True (#19766)

BowenBao committed 5 years ago

831bd1c2

Make graph->param_node()->next() the first node (#19788)

zdevito committed 5 years ago

e58817fe

Enable ROCm multi-gpu with Gloo

bddppq committed 5 years ago

bc539845

Add proper __repr__ to LogSoftMax (#20018)

sdg committed 5 years ago

cf55670b

Working on component-wise transformations that mimic `torch.cat` and `torch.stack` (#11868)

ahmadsalim committed 5 years ago

b8256280

add process_group in convert_sync_batchnorm (#19240)

zhangliliang committed 5 years ago

f7a78688

Add DirichletFullyConnectedActor to Soft Actor-Critic

econti committed 5 years ago

2356fac9

Add Custom graph fusion (#18588)

bwasti committed 5 years ago

4ca325df

Intra-op parallel microbenchmarks for PT (#19997)

Ilia Cherniavskii committed 5 years ago

19e68865

Allow a non-OpenMP based build (#19749)

Ilia Cherniavskii committed 5 years ago

481b6d02

Newer Older