strisunshinewentingwang/pytorch

Pull Requests Commits

Add test cases for optional of list (#19592)

smessmer committed 5 years ago

fddd763e

Port adaptive_max_pool3d() to ATen (#19547)

skrah committed 5 years ago

fc8834df

add torch.tensor requires grad (#19445)

Elias Ellison committed 5 years ago

0922a64d

Surface the Glow traces to C2 (#19087)

Yinghai Lu committed 5 years ago

4e8cc8ee

Fix lack of state init for adagrad and add share_memory flag (#17679)

Stonesjtu committed 5 years ago

444f792f

Allow extracting element-wise loss in softmax (#19579)

prigoyal committed 5 years ago

0d0acba3

dispatch max_pools with no indices, expose max_pools to torch namespace (#19449)

wanchaol committed 5 years ago

e9c8f372

Adds `fakeQuantizePerTensorAffineOp` to pytorch (#19387)

jerryzh168 committed 5 years ago

f3be2816

-fno-math-errno -fno-trapping-math (#19552)

James Reed committed 5 years ago

1b3967b4

Only require python print on certain namespaces (#19383)

bwasti committed 5 years ago

d8729efa

Use `fbgemm` for quantize/dequantize ops (#19500)

Zafar Takhirov committed 5 years ago

3cc60e54

Specify to use Float16UniformFill if necessary in sparse lookup layer (#18499)

chocjy committed 5 years ago

714344a9

Fix the Division by Zero Bug of CosineAnnealingLR (#19180)

chandlerzuo committed 5 years ago

e3f15046

Fix the documentation for BCEWithLogitsLoss (#17218, #16804) (#19212)

vvelicodnii-sc committed 5 years ago

7a418969

fix the docstring of `RandomSampler` (#19113)

crcrpar committed 5 years ago

bb05f707

Avoid (future) cusparse name collision (#19591)

mruberry committed 5 years ago

83cf9473

Add docs and test guaranteeing indices from torch.nonzero ordered C-style (#19539)

jhultman committed 5 years ago

f767c9ac

Remove unnecessary printing from tests

ssnl committed 5 years ago

3b4d4ef5

Fix lr_scheduler's last_epoch value at the time of initialization (BC BREAKING!) (#7889)

bado-lee committed 5 years ago

36084908

Removes variable which is assigned but not used (#19194)

SebFar committed 5 years ago

f9c4ce78

add torch.cuda.synchronize(device=None) (#19573)

ssnl committed 5 years ago

dce3d74d

Port adaptive_max_pool2d() to ATen (#19409)

skrah committed 5 years ago

75ce5173

Fix math formatting of PairwiseDistance and CosineSimilarity docs and fix math formatting of CTC loss docs.

zhiqwang committed 5 years ago

88f78c71

Revert D15039713: [pytorch][PR] add torch.tensor requires grad

suo committed 5 years ago

5bafb64e

Bugfix for fusion device check (#19594)

James Reed committed 5 years ago

e7fc7c73

add torch.tensor requires grad (#19445)

Elias Ellison committed 5 years ago

d2b03512

Add onnx support for _unique2 operator

VitalyFedyunin committed 5 years ago

8be6d5ff

Automatic update of fbcode/onnx to 0e8d2bc5e51455c70ef790b9f65aa632ed9bc8a7 (#19568)

houseroad committed 5 years ago

5a796d15

Don't create FusionGroups for known-CPU producer values (#19342)

James Reed committed 5 years ago

5be4bee4

Explicitly define supported types (#19516)

smessmer committed 5 years ago

969af431

Newer Older