strisunshinewentingwang/pytorch

Pull Requests Commits

[JIT] during freezing, cast optional bias to half if weight is half

davidberard98 committed 2 years ago

2547be51

Support comparing Sparse Compressed tensors

pearu committed 2 years ago

9a608af8

[GH1] Allow conclusions to include workflows and checks (#77597)

janeyx99 committed 2 years ago

f9aaf9d3

[shard] add clone/detach and set requires_grad for ShardedTensor

wanchaol committed 2 years ago

25fa964d

Revert "Revert "[LT] Codegen ReuseNode for supported ops""

desertfire committed 2 years ago

25c6ebd1

check that flip doesn't accept repeating dimensions (#77500)

ngimel committed 2 years ago

375e21b2

Use FastAtomicAdd for index_add/_index_reduce

mikaylagawarecki committed 2 years ago

fe1232a4

[mergebot cli] add new revert regex (#77459)

zengk95 committed 2 years ago

906fc777

Add Sparse CSC support to torch.empty

pearu committed 2 years ago

d76efed5

Replace SparseCsrTensorImpl crow/col_indices with compressed/plain_indices

pearu committed 2 years ago

77f1ca00

stirde_properties fix (#77460)

jjsjann123 committed 2 years ago

1b1dec45

Bugfix gradient formula for index_reduce('prod') + separate out sample_inputs for index_reduce

mikaylagawarecki committed 2 years ago

7ba4e124

Compressed sparse layout conversion stubs (#77489)

cpuhrsch committed 2 years ago

8c608a79

Revert "enable NVFuser by default"

pytorchmergebot committed 2 years ago

2a905aef

skip primTorch nvfuser tests on rocm (#77468)

jjsjann123 committed 2 years ago

8875453d

[Model Averaging] Support disabling post-local gradient sync (#76723)

wayi1 committed 2 years ago

5ab8afe4

[coreml] Use special throw macro when encountering CoreML API errors (#77429)

SS-JIA committed 2 years ago

93d84c0f

[metal] Use special throw macro when encountering Metal API errors (#77430)

SS-JIA committed 2 years ago

c34be4eb

[GHA] Use silent checkout (#77547)

malfet committed 2 years ago

46e16737

[chalf] exp, log: cuda (#77483)

kshitij12345 committed 2 years ago

a8261b4b

[metal] fix op registration for cat (#77451)

SS-JIA committed 2 years ago

9c94b760

Adding maximize to Adamax (#77409)

zaxtax committed 2 years ago

2a496e2f

[Profiler] Remove `addMemoryUsageActivity` from kineto_shim (#76678)

Taylor Robie committed 2 years ago

c3c92ca1

[BE][CI] Add `pip_install` macro

malfet committed 2 years ago

1db13374

Python function to extract information on mobile::Module from flatbuffer (#77328)

qihqi committed 2 years ago

69fa49f1

Fix optional dtype/layout/memory_format pycall; fix memory format

ezyang committed 2 years ago

b5bc954a

Saving JIT to flatbuffer should respect options. (#77456)

qihqi committed 2 years ago

14e59edd

Disable GC if fraction is not set (#76648)

jaewonlee-fb committed 2 years ago

3e89a1d6

add BFloat16 support for BatchNorm on CPU

mingfeima committed 2 years ago

8c504142

Revert "[mobile] add buck build for mobile targets (#76480)"

pytorchmergebot committed 2 years ago

530481ed

Newer Older