strisunshinewentingwang/pytorch

Pull Requests Commits

quantization: improve documentation on natively supported backends (#58925)

vkuzo committed 3 years ago

dafa4b35

[Reland2][DDP] Merge work and future_work in reducer (#59574)

Yi Wang committed 3 years ago

6575975d

Use irange in torch/csrc/jit (#55716)

r-barnes committed 3 years ago

fbe65b16

enable upload test stats on PR (#59567)

walterddr committed 3 years ago

ff553e5b

Adding run-specified-test-cases option in run_test.py (#59487)

janeyx99 committed 3 years ago

24432eaa

Move sharding to after all tests have been excluded (#59583)

janeyx99 committed 3 years ago

caf76c24

Use irange in a few places (#55325)

r-barnes committed 3 years ago

93140a31

Strictly type everything in .github and tools (#59117)

samestep committed 3 years ago

737d920b

DOC Improve documentation for LayerNorm (#59178)

thomasjpfan committed 3 years ago

6ff001c1

fix double backward for `binary_cross_entropy` loss function when `reduction=sum`. (#59479)

nikitaved committed 3 years ago

a30b3595

Fix error message formatting in _make_grads (#59532)

take-cheeze committed 3 years ago

77dde35f

[ROCm] enable kernel asserts (#49624)

jeffdaily committed 3 years ago

24e27af6

fix name of 'dims' kwarg in torch.tile docs (#59471)

saketh-are committed 3 years ago

05b571ee

Add warning about should_drop for JIT coverage plug-in (#57961)

janeyx99 committed 3 years ago

b0ac9bfb

DOC Small rewrite of interpolate recompute_scale_factor docstring (#58989)

thomasjpfan committed 3 years ago

8693e288

[PyTorch] Optimize c10::optional<ArrayRef<T>> for size (#59333)

swolchok committed 3 years ago

1798ff02

[quant] Implemented InputWeightObserver for Linear inputs

angelayi committed 3 years ago

cc03ea2c

Make `binary_cross_entropy` differentiable wrt `target` (#59447)

nikitaved committed 3 years ago

c51abf8f

Revert D28922305: [Reland][DDP] Merge work and future_work in reducer

Mike Ruberry committed 3 years ago

94cc681f

Revert D28922548: [Gradient Compression] Apply division first to avoid overflow

Mike Ruberry committed 3 years ago

f998e63d

[Gradient Compression] Apply division first to avoid overflow (#59522)

Yi Wang committed 3 years ago

459270ac

Adding users of a node to the serialized JSON. (#59357)

Protonu Basu committed 3 years ago

a2e56fa0

Adds remaining OpInfos and removes redundant test generators (#55558)

Mike Ruberry committed 3 years ago

de40c8e4

[PyTorch Edge] Remove legacy and kineto profilers from mobile build (#58730)

dhruvbird committed 3 years ago

8c852de5

[Reland][DDP] Merge work and future_work in reducer (#59520)

Yi Wang committed 3 years ago

3137bbeb

Migrate `torch.lstsq` to ATen (#59400)

peterbell10 committed 3 years ago

390fe749

OpInfo: to_sparse (#59445)

kshitij12345 committed 3 years ago

da972afd

OpInfo: t (#59442)

kshitij12345 committed 3 years ago

96ac0e03

Support `__rmod__` (#58476)

asi1024 committed 3 years ago

0a5bfa99

flip via TI (#59509)

ngimel committed 3 years ago

344ecb2e

Newer Older