strisunshinewentingwang/pytorch

Pull Requests Commits

Port `remainder` from TH to ATen (CPU and CUDA) (#34136)

kurtamohler committed 4 years ago

fbbeee09

[pytorch] remove boilerplate setQEngine() from PyTorch mobile predictors (#34556)

ljk53 committed 4 years ago

7aca9afd

AccumulateGrad: ensure sparse tensor indices and values refcount is always 1 (#34559)

pritamdamania committed 4 years ago

2ce9513b

Add Tensor overload for start in narrow. (#34317)

Ailing Zhang committed 4 years ago

ab2297df

add quantized_hardtanh (#34097)

vkuzo committed 4 years ago

2e88a78d

Fix static data initialization deadlock on GIL (#34505)

xush6528 committed 4 years ago

8d84c5f1

Set USE_RCCL cmake option (dependent on USE_NCCL) (#31341)

jithunnair-amd committed 4 years ago

ce77d4a3

[jit] Add type tags to lists/dicts in pickle (#33255)

davidriazati committed 4 years ago

23b2fba7

[pytorch][ci] add build_only flag to mobile CI jobs (#34560)

ljk53 committed 4 years ago

4167db11

[pt][quant] Vectorized qmul and more methods on qint data types (#34376)

dskhudia committed 4 years ago

a09c4d39

[JIT] Introduce a fake Tensor creation node for IR unit tests (#34334)

Meghan Lele committed 4 years ago

903ad903

Preserve memory format for torch.cat on CUDA (#34526)

zasdfgbnm committed 4 years ago

d0834c5b

convert counter back to list #33229 (#33356)

vincentqb committed 4 years ago

be3bc1de

Do not use clang if it can not parse system extensions (#34549)

malfet committed 4 years ago

dd7cec68

Add the build for runtime dispatch for AVX, AVX2 instruction set (#26125)

lly-zero-one committed 4 years ago

09296c34

[caffe2] do not declare __assert_fail in clang builds (#33893)

Igor Sugak committed 4 years ago

259d7299

[C++ API Parity] rmsprop optimizer update (#33450)

anjali411 committed 4 years ago

2d24005d

Change std::to_string call to c10::to_string

dreiss committed 4 years ago

6f12145c

Turn on exact_dtype by default on test_sparse.py (#34489) (#34542)

terhuhf committed 4 years ago

2cf344be

Avoid clone for sparse tensors during accumulation of grads. (#33427)

pritamdamania committed 4 years ago

b185359f

.circleci: Switch should_run_job cuda 10.1 -> 10.2 (#34498)

seemethere committed 4 years ago

5f61f42c

fix handling of replica parameters in DataParallel (#33907)

ngimel committed 4 years ago

cd9d9a22

Clean up include list of Shape.cu (#34528)

zasdfgbnm committed 4 years ago

0dbfb26e

remove duplicated process group gloo timeout (#31342)

zhaojuanmao committed 4 years ago

cb689a5d

Fix #33562 (uncaught domain_error on macOS) (#34301)

andrewdelong committed 4 years ago

c7dd5f89

Check if rnn weights need to be flattened (#34265)

jithunnair-amd committed 4 years ago

9e94e464

[ROCm] Enable BFloat16 type for loss functions and few misc ops required for resnet50 (#34469)

rohithkrn committed 4 years ago

29b67339

Update compiler warning about ABI compatibility (#34472)

ppwwyyxx committed 4 years ago

20b18a58

Remove custom function in no_grad block error message (#33896)

zou3519 committed 4 years ago

f5ee46f1

Print the current Node name in anomaly mode (#33875)

zou3519 committed 4 years ago

3e6e2e9b

Newer Older