strisunshinewentingwang/pytorch

Pull Requests Commits

Set stream for softmax kernel launch (#22470)

Kexin Yu committed 5 years ago

e5d64034

update video input (#22471)

Du Tran committed 5 years ago

d2ceab27

Add the support of handle Bias being nullptr for torch.ops.quantized.fbgemm_conv (#22472)

jianyuh committed 5 years ago

4ba1c4f7

suo committed 5 years ago

57dbc796

use save/load for emitFunctionHook

suo committed 5 years ago

b952011b

remove unused argument in import.cpp (#22205)

suo committed 5 years ago

bc24593a

improvements to QualifiedName (#22204)

suo committed 5 years ago

4b9b7d6f

refactoring of module/object (#22203)

suo committed 5 years ago

f5919dba

Make CompilationUnit own Functions (#22202)

suo committed 5 years ago

3b2844ee

make test context managers exception safe

suo committed 5 years ago

66378c70

add #pragma once to jit headers

Krovatkin committed 5 years ago

2b06e7cd

remove erase_fork_wait.h

Krovatkin committed 5 years ago

6f6a6804

move casting ops from prim to aten

wanchaol committed 5 years ago

799633e4

Rereapply optional ScalarType interface changes that were reverted in D16079809 (#22456)

nairbv committed 5 years ago

97a604ef

Remove weak script (#22212)

David Riazati committed 5 years ago

10c4b98a

add JIT path to the benchmark (#22309)

mingzhe09088 committed 5 years ago

b93f29de

Fix SyncBatchNorm running var update issue (#22248)

unlimblue committed 5 years ago

29ec4769

create tensor based on provided datatype (#22468)

mingzhe09088 committed 5 years ago

325ec232

Fix onnx custom op export & add initial test case (#21321)

BowenBao committed 5 years ago

319ef3bc

generate tests based on op metadata (#21432)

mingzhe09088 committed 5 years ago

9c44f6c7

Another dce fix (#22499)

smessmer committed 5 years ago

2732a5e5

Perform weight re-init for embedding table in sparse_lookup.py (#22348)

Alyssa Wang committed 5 years ago

d9e15bcc

Add device guard around MPI operations (#22446)

pietern committed 5 years ago

c9f41e9b

Don't construct a single element array for unary ops

James Reed committed 5 years ago

abb2e689

Take const refs in TensorIterator::mark_outputs

James Reed committed 5 years ago

7fef0b7b

Deprecate vector/unordered_map again (#22478)

smessmer committed 5 years ago

0d636194

Fix dead code elimination in onnx export (#22476)

smessmer committed 5 years ago

17cc7986

remove caffe2/core dependency from ATen/core/jit_type.h (#22441)

ljk53 committed 5 years ago

76e14c1e

Add `torch.where` overload with only condition argument (#21986)

Brennan Vincent committed 5 years ago

e210c650

provide "size" parameter in torch.normal when called with two floats (#20545)

Brennan Vincent committed 5 years ago

dcd902bd

Newer Older