strisunshinewentingwang/pytorch

Pull Requests Commits

Move legacy Caffe2 TensorImpl methods out of header

ezyang committed 2 years ago

4bd5b161

[WIP] Make constructor calls in experimental MetaTracer serializable

James Reed committed 2 years ago

7311390d

add channels last support for thnn_conv2d (non-dilated)

mingfeima committed 2 years ago

3d561ee9

Fix Optimizer.zero_grad type annotation (#76998)

James Reed committed 2 years ago

57b54dfe

[shard] add more tensor creation ops (#77185)

wanchaol committed 2 years ago

81528d4b

`nn.functional.glu`: forward AD support (#77186)

nikitaved committed 2 years ago

afd8bd77

Fix for normalizing signature for op overloads (#77182)

Elias Ellison committed 2 years ago

023aafbc

[SR] Fuse clamp/nan_to_num

mikeiovine committed 2 years ago

02713221

Use a more reliable signaling mechansim to stop TCPStore background threads (#76973)

yifuwang committed 2 years ago

27ea79b8

Revert "[ci] delete old linter stuff"

pytorchmergebot committed 2 years ago

12f67ed2

ci: Move rocm distributed tests to periodic

seemethere committed 2 years ago

872f5daf

Add `load_state_dict` error message for non-dicts (#77197)

neverix committed 2 years ago

87e543da

[ci] delete old linter stuff

suo committed 2 years ago

3a68155c

Fix case of PyTorch in issue templates (#77180)

daniellepintz committed 2 years ago

ec33f0ee

Fix bincount to use acc scalar for the bounds (#76979)

t-vi committed 2 years ago

58bb1f74

Jiterator with Python Registration (#77121)

SherlockNoMad committed 2 years ago

61dcde88

[chalf] update type promotion table (#76893)

kshitij12345 committed 2 years ago

00fb8282

Move `contiguous_strides` to `c10/util/strides.h`.

ysiraichi committed 2 years ago

2e6ed5d0

replace "grep" with Python processing in `collect_env.py` (#77148)

pmeier committed 2 years ago

635aaa3d

torch/deploy,package: log usage for InterpreterManager, PackageExporter, PackageImporter (#77097)

d4l3k committed 2 years ago

4d30ebc8

[ROCm] update cmake package DIR paths (#77087)

jeffdaily committed 2 years ago

64b54343

shard `linux-xenial-py3.7-clang7-asan / test (default` from 4 to 5 (#77084)

clee2000 committed 2 years ago

ecab4fb0

Revert "speed up 1d sort (#77100)"

pytorchmergebot committed 2 years ago

54dbfc1e

faster batch sampler (#76951)

木澄 committed 2 years ago

d22d749a

Fix `broadcast_in_dim` support in NVFuser Frontend (#76790)

kevinstephano committed 2 years ago

752d496c

Parallelize `gatherTopK` on multiple blocks (#74267)

zasdfgbnm committed 2 years ago

8560fa73

Add meta tensor support for some operations using python registration

anjali411 committed 2 years ago

767af8e3

Profiling range for FSDP.forward (#76899)

rohan-varma committed 2 years ago

e31b6213

Fix max pool forward nhwc (#76597)

Aidyn-A committed 2 years ago

a127c584

Add __all__ for torch.autograd.{anomaly_mode, gradcheck, forward_ad}

anjali411 committed 2 years ago

65a8f8f6

Newer Older