strisunshinewentingwang/pytorch

Pull Requests Commits

Make build with makefiles less noisy (#21053)

bddppq committed 5 years ago

5c53aa48

Fix get_gpu_memory_info in non-cuda builds (#21054)

bddppq committed 5 years ago

9b147961

Deprecate variadic inputs of checkpoint_sequential (#21006)

sublee committed 5 years ago

ffdce790

Allow nondet_tol for nondeterminism in gradcheck and gradgradcheck (#20980)

t-vi committed 5 years ago

d23d04f1

Fix typo in CyclicLR docs (#21021)

njdalton committed 5 years ago

d190450a

add simple memory analyzer and log warning if GPU underutilized (#21024)

Aapo Kyrola committed 5 years ago

f1fe4b11

Fix warning in register_c10_ops by making index unsigned (#20964)

Chillee committed 5 years ago

1bed5f39

Enable batched QR decomposition and add a `some` option (#20689)

vishwakftw committed 5 years ago

f6ec4648

Use constants from math.h for gelu op (#20974)

xiaomengy committed 5 years ago

c1048182

tracing for intra_op_parallel (#20603)

jspark1105 committed 5 years ago

0290897b

Add an option to stop the build process once cmake terminates. (#21034)

xuhdev committed 5 years ago

9a989ec4

Add `Tensor.T` attribute to reverse dimensions (#20598)

Brennan Vincent committed 5 years ago

9294de8c

Renaming the relu kernel and adding hypothesis tests (#20647)

Zafar Takhirov committed 5 years ago

2791a449

Added engine information to the profiling result. (#20493)

kimishpatel committed 5 years ago

d6d192e0

Enable operator profiling via command line (#20173)

kimishpatel committed 5 years ago

7afa7500

Fixed gcd to use 64 bit integers (#21041)

Chillee committed 5 years ago

2ba608b4

Make ScriptModule.training an attribute instead of a parameter (#19587)

David Riazati committed 5 years ago

28079c39

Better error message in __get_state__ to let a user know that ScriptModules can't be deep-copied atm (#20885)

Krovatkin committed 5 years ago

18809f7b

Some minor fixes for the changes in #20945 (#21008)

peterjc123 committed 5 years ago

07c4e45c

refactor register_prim_ops (#21001)

wanchaol committed 5 years ago

0885dd28

Re-land "Fix advanced indexing on "huge" Tensors" (#21019)

colesbury committed 5 years ago

b85c5292

Improve error message for missing attribute (#20779)

Chillee committed 5 years ago

52d27890

Some name and variable cleanup (#20861)

orionr committed 5 years ago

bc10677f

Re-enable test_dag_net_forking on ROCm (#21013)

bddppq committed 5 years ago

99674eb8

Clarify cycliclr param docs (#20880)

sampepose committed 5 years ago

082936f0

Change bound shape inference for LengthsRangeFill & GatherRanges, add more tests (#20610)

ChunliF committed 5 years ago

68c3ef72

Refactor schema_matching.cpp (#20549)

davidriazati committed 5 years ago

bbe34118

Generate TH functions outside of Type (#20309)

Roy Li committed 5 years ago

ff6cda0d

Move 1d tensor checks to TH (#20859)

Roy Li committed 5 years ago

eacb3118

Change view dispatch to abstract (#20308)

Roy Li committed 5 years ago

d2f14db6

Newer Older