strisunshinewentingwang/pytorch

Pull Requests Commits

[C++ API] Update torch::nn layer docs (#34522)

Will Feng committed 4 years ago

9064fafb

[JIT] Add support for tolist for GPU-resident Tensors (#34554)

Meghan Lele committed 4 years ago

56832bf7

[ci] try to fix rocm builds (#34600)

suo committed 4 years ago

866505b1

Fix typo in documentation (#34581)

kvthr committed 4 years ago

2de4f245

[On-device Benchmark] speed_benchmark_torch switch to log latency from dataset level to row level (#34598)

Peng Xia committed 4 years ago

25e4e9eb

Fix SELECTED_OP_LIST file path issue (#33942)

gemfield committed 4 years ago

70f32986

[JIT] Torchbind error if python instantiate class that doesnt exist (#34568)

James Reed committed 4 years ago

1f834b5c

Disable ROCM when building mobile libtorch. (#34478)

syoyo committed 4 years ago

12fb8148

[distributed] quicker exit in the case of failed tests in distributed (#34150)

rohan-varma committed 4 years ago

b553e691

small typos (#34589)

ettiee committed 4 years ago

2cf576e9

Stop last usage of newWithSize. (#34386)

gchanan committed 4 years ago

82cdd3ab

Revert D20193196: [pytorch][PR] PCA and SVD for low-rank matrices, LOBPCG for positive-defined generalized eigenvalue problem

ezyang committed 4 years ago

4b929e54

Revert D20282846: Delete OperatorOptions, absorb AliasAnalysisKind into FunctionSchema.

ezyang committed 4 years ago

6f8a8e4e

Revert D20379910: [pytorch][PR] Set USE_RCCL cmake option (dependent on USE_NCCL)

ezyang committed 4 years ago

63964175

PCA and SVD for low-rank matrices, LOBPCG for positive-defined generalized eigenvalue problem (#29488)

pearu committed 4 years ago

2ec779d4

Stop using ctypes to interface with CUDA libraries. (#33678)

peterbell10 committed 4 years ago

5fc5cf65

Delete OperatorOptions, absorb AliasAnalysisKind into FunctionSchema. (#34160)

ezyang committed 4 years ago

9d42177a

Beef up documentation on Dispatcher.h, reorder methods for clarity. (#33838)

ezyang committed 4 years ago

b2344b70

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

Newer Older