strisunshinewentingwang/pytorch

Pull Requests Commits

remove scalar to float matching (#19918)

Elias Ellison committed 5 years ago

ccf35f4b

Misc pickler improvements (#19638)

davidriazati committed 5 years ago

4294dba9

Remove self-include. (#19833)

Mikhail Zolotukhin committed 5 years ago

8933ff65

Enabled masked for a bool tensor (#19140)

izdeby committed 5 years ago

de19eeee

Add magma for CUDA 10.1 to Windows docs

peterjc123 committed 5 years ago

39b885cb

Build caffe2/fb/operators (#19688)

xw285cornell committed 5 years ago

9d0b5a1c

Finer grained consistency check in reducer (#19901)

pietern committed 5 years ago

84136002

Only call into reducer if torch.is_grad_enabled() (#19897)

pietern committed 5 years ago

5525c419

Remove trailing whitespace flake8 lint (#19828)

suo committed 5 years ago

a92e1ccd

Make find_unused_parameters in DDP default to False (#19895)

mrshenli committed 5 years ago

b695e562

Remove redundant include from jit/fuser/cpu/dynamic_library.h. (#19863)

Mikhail Zolotukhin committed 5 years ago

3f7a8778

Fix conda build for Windows (#19824)

peterjc123 committed 5 years ago

3803d1c9

Allow for iterations where no module parameter is used (#19821)

pietern committed 5 years ago

9b69da2b

Use QualifiedName for classes (#19575)

suo committed 5 years ago

f0a007a2

make QualifiedName a value type (#19626)

suo committed 5 years ago

1d6e868c

separate QualifiedName into its own file (#19566)

suo committed 5 years ago

096dd8a4

Avoid Output Uninitialized Blobs in Load with load_all=1 (#19133)

chandlerzuo committed 5 years ago

472be69a

Fix CUDA stream syncing bug in allgather and reduce_scatter (#19631)

mxw committed 5 years ago

268859ce

use fully qualified name for ScriptClasses (#19239)

suo committed 5 years ago

a25b7953

Add elementwise_affine for layer_norm_op (#19713)

xiaomengy committed 5 years ago

2ce39de3

Add support for LONG_BINGET pickler op (#19815)

David Riazati committed 5 years ago

f9786ad3

fix optional type unification (#19813)

Elias Ellison committed 5 years ago

5a83a742

DataLoader docs update to describe how workers are managed, including Windows. (#18091)

mike9ant committed 5 years ago

698103cd

Revert D15103223: [pytorch][PR] [CUDA 10] Resolve host_define.h warnings

soumith committed 5 years ago

4e6608e8

update F.grid_sample doc for clarity (#19754)

ssnl committed 5 years ago

42fbeef5

Cleanup documentation (#19584)

davidriazati committed 5 years ago

dc67d9f3

Revert D14577575: [pytorch][PR] Fix lack of state init for adagrad and add share_memory flag

soumith committed 5 years ago

75754bec

Fix the install of TensorBoard for doc generation (#19814)

orionr committed 5 years ago

11297702

Follow up to adaptive_max_pool3d() port (#19748)

skrah committed 5 years ago

be20d65b

Follow up to adaptive_max_pool2d() port (#19738)

skrah committed 5 years ago

cb4d41af

Newer Older