strisunshinewentingwang/pytorch

Pull Requests Commits

[nnc] Add call_with_numel interface for fast CUDA calls (#65213)

bertmaher committed 2 years ago

8cf047af

Avoid UB when indexing into size-0 tensors (#65878)

bertmaher committed 2 years ago

8595b6ee

Improve pytorch type hints (Dataloader, trig functions)

shapovalov committed 2 years ago

fc52f129

[Static Runtime] Fuse ListUnpack + gather_ranges_to_dense (#65116)

Mike Iovine committed 2 years ago

982ef883

pytorch quantization ao migration phase 2: caffe2/test (#65832)

vkuzo committed 2 years ago

227e37dd

pytorch quantization ao migration phase 2: torch/jit (#65829)

vkuzo committed 2 years ago

dac35b35

pytorch quantization ao migration phase 2: caffe2/benchmark (#65833)

vkuzo committed 2 years ago

e3af4be9

[special] special alias for softmax (#62251)

kshitij12345 committed 2 years ago

c1447f06

[sparsity] Add m-out-of-n support in the WeightNormSparsifier (#65295)

z-a-f committed 2 years ago

c27b427c

[sparsity] Change API to take FQNs as configuration (#65296)

z-a-f committed 2 years ago

8b1aa853

[PyTorch] Avoid string construction from const char* and speedup empty string creation if error messages are suppressed (#65939)

dhruvbird committed 2 years ago

ea0de37d

Added jit log stream changing function and some refactor (#65768)

narang99 committed 2 years ago

2828ce53

[deploy][1/n] Make deploy code conform to PyTorch style. (#65861)

suo committed 2 years ago

33c03cb6

[TensorExpr] Move lowerings registration from kernel.cpp to lowerings.cpp. (#65553)

Mikhail Zolotukhin committed 2 years ago

765b6a90

[TensorExpr] Move 'compute*' functions to operators/... (#65552)

Mikhail Zolotukhin committed 2 years ago

015e0079

[TensorExpr] Port NNC lowerings to the new registry mechanism. (#65551)

Mikhail Zolotukhin committed 2 years ago

3a0165da

[TensorExpr] Add a skeleton for a registry of NNC lowerings. (#65550)

Mikhail Zolotukhin committed 2 years ago

eee9ad0f

[TensorExpr] Kernel: make prim::ConstantChunk handled like other ops. (#65549)

Mikhail Zolotukhin committed 2 years ago

d84191fc

[Static Runtime] Make module_ optional in StaticModule (#65882)

Hao Lu committed 2 years ago

a6ad2b41

slow_conv2d grad_input: avoid dispatch in parallel region (#65725)

peterbell10 committed 2 years ago

08df4c2b

Make JIT Aliasing Test Less Brittle (#65493)

Jiewen Tan committed 2 years ago

6502fb89

[QPL] move metadata logging to markerEnd for model run QPL (#65451)

linbinyu committed 2 years ago

4f5ea598

[caffe2] fix LLVM-12 nullptr-with-nonzero-offset UBSAN error (#65506)

Igor Sugak committed 2 years ago

2481c064

Migrate THCCachingHostAllocator to ATen (#65746)

peterbell10 committed 2 years ago

f6dfac69

[ONNX] Enable export of __xor_ (#64042) (#64581)

BowenBao committed 2 years ago

d3979034

[ONNX] Fix inplace fill_ dtype export mismatch (#64233) (#64580)

BowenBao committed 2 years ago

e598ba2e

[ONNX] Update doc and error message for indexing export (#64290) (#64579)

BowenBao committed 2 years ago

89cbe622

[ONNX] Fix remainder export (#64230) (#64578)

BowenBao committed 2 years ago

d4ff344f

Add onnx test for batched_nms (#53175) (#64381)

BowenBao committed 2 years ago

0f0ef4fe

[ONNX] Fix gather squeeze axis in constant folding (#63588) (#64379)

BowenBao committed 2 years ago

7e15f2dd

Newer Older