strisunshinewentingwang/pytorch

Pull Requests Commits

Update deprecated import paths. (#67250)

datumbox committed 2 years ago

38cbaeb8

[Static Runtime] Add more debug info to verify_no_memory_overlap() (#67206)

Hao Lu committed 2 years ago

0c1b7545

[sharded_tensor] refactor sharded_tensor file structure (#67199)

wanchaol committed 2 years ago

31bcfa37

add frozen_pyyaml as a builtin library to torch::deploy (#67127)

zyan0 committed 2 years ago

b96337cf

[fx-acc] add automated graph opt testing using AccOpProperty (#67228)

alexbeloi committed 2 years ago

0e371e41

Add torch.nn.init.normal_ and torch.nn.init.kaiming_uniform_ ops to ShardedTensor (#67057)

bowangbj committed 2 years ago

3596e13d

[trt] Algorithm recorder/replayer (#4)

Yinghai Lu committed 2 years ago

bfcde086

[Light] Remove ambiguity from compile_spec names, use actual output type (#67209)

priyaramani committed 2 years ago

ecf7e969

[PyTorch] Add flop count for bmm and baddbmm (#66636)

hjmshi committed 2 years ago

ad5731ca

[PyTorch Edge][type] Add type support for NamedTuple custom class (export) (#62612)

cccclai committed 2 years ago

7acf0c6d

Disallow annotations on instance attributes outside __init__ (#67051)

andrewor committed 2 years ago

0d7d4461

[WIP] wrap XLATensors into Python XLA wrapper class (#65841)

Krovatkin committed 2 years ago

1f55dd83

[skip ci] Set test owner for test_profiler.py (#66831)

janeyx99 committed 2 years ago

fa7fb7b4

[vulkan] Add 2D transposed convolutions (#67104)

SS-JIA committed 2 years ago

0acc21b4

[jit] Factor findAllNodes into one place. (#65965)

zhxchen17 committed 2 years ago

059ae960

[fx2trt] Better trt layer name (#67200)

842974287 committed 2 years ago

239b3826

[quant] Remove unused print in quantization_patterns.py (#67191)

jerryzh168 committed 2 years ago

4ac8d069

[jit][edge] Enable CALL instruction in lite interpreter. (#65964)

zhxchen17 committed 2 years ago

12daa4f6

Refactor cub namespace handling (#66219)

zasdfgbnm committed 2 years ago

b8dfb45a

Sparse CSR CUDA: add `torch.addmm` with all inputs sparse (#63511)

IvanYashchuk committed 2 years ago

700b39a3

Improve assert failure message in test_get_torch_func_signature_exhaustive (#67039)

pearu committed 2 years ago

333717ea

[Pytorch Edge] Extend runtime compatibility to custom classes (#66972)

JacobSzwejbka committed 2 years ago

a6d03394

Better and more consistent error messages in torch.linalg (#62734)

lezcano committed 2 years ago

f4dd8848

[jit][edge] Add control stack frame to lite interpreter (#65963)

zhxchen17 committed 2 years ago

4dce051c

.github: Migrate linux-xenial-py3.6-gcc7 to GHA (#67072)

zhaoalex committed 2 years ago

ac948f4f

Move the registration of CPython builtin modules to BuiltinRegistry (#67085)

PaliC committed 2 years ago

9de08888

Disable SVE when cross-compiling for M1 (#67114)

malfet committed 2 years ago

d68bb50e

[Static Runtime] Add static_runtime::fused_sigrid_transforms (#66659)

Mike Iovine committed 2 years ago

5d9ff8f3

Use `at::native::is_nonzero` in promoted ops to improve portability (#67097)

pavithranrao committed 2 years ago

8d164a36

[Pytorch][Bootcamp] Add fixes and vanilla testing for Adagrad non-vectorized and vectorized optimizers to handle complex numbers (#66671)

Christopher Gray Howard committed 2 years ago

acb340de

Newer Older