strisunshinewentingwang/pytorch

Pull Requests Commits

Resolve with closed over variables instead of stack frame (#22270)

davidriazati committed 5 years ago

2891784a

Fix load op to return the shape info as before when loading multiple blobs (#23166)

pradeepdfb committed 5 years ago

fd90b967

Added running via throughput benchmark options. (#23077)

kimishpatel committed 5 years ago

82db5dce

Added add op framework overhead benchmark for C2 (#23078)

kimishpatel committed 5 years ago

2ba516d5

Add simple add op based framework overhead benchmark. (#23076)

kimishpatel committed 5 years ago

0621068c

fix qat tests (#23124)

jerryzh168 committed 5 years ago

4223e2f9

Remove cuda free mutex (#23040)

jeffdaily committed 5 years ago

8bc28cc8

izdeby committed 5 years ago

22f7c9e3

Revert "Temporarily skip mypy-0.720 to unbreak master type checks"

zou3519 committed 5 years ago

aeee49d5

Update ScatterWeightedSum Op (#23087)

rohithkrn committed 5 years ago

b8c8977b

hipify: do not overwrite files that stay the same (#23112)

t-vi committed 5 years ago

ff8cb9f6

Fix memory leak in Adam, Adagrad, RMSProp (#23125)

jerry73204 committed 5 years ago

2ac9abf7

improve enforce in cross_entroy_op (#23062)

Wakeupbuddy committed 5 years ago

96b6797f

z-a-f committed 5 years ago

3e663850

MaxPool2d in the torch (#22765)

Zafar Takhirov committed 5 years ago

963707c5

Concat with out (#22408)

Zafar Takhirov committed 5 years ago

cf3e6478

make jit logging visible, so it can be used in a TVM compiler

Krovatkin committed 5 years ago

05f088ec

Use set -x to help investigate doc push errors (#23111)

kostmo committed 5 years ago

bb9119f6

Remove unused atomics detection code. (#23089)

xuhdev committed 5 years ago

a62c6874

fix indexing for more than 65535 elems in non-indexed first dim (#23123)

ngimel committed 5 years ago

4e5f7008

support at::view and at::reshape for quantized tensor (#23046)

jerryzh168 committed 5 years ago

6791f395

Move THTensor_compute_stride to ATen (#23045)

jerryzh168 committed 5 years ago

a03205ed

Add fused modules in nn._intrinsic (#23085)

jerryzh168 committed 5 years ago

0d8324b1

Quantized concatenation (+fused relu). (#21749)

Zafar Takhirov committed 5 years ago

47af41fe

Moving np function to test area

z-a-f committed 5 years ago

9f4df63c

Conv module (#23084)

jerryzh168 committed 5 years ago

77353636

Fbgemm fp16 tensor support (#23101)

Yinghai Lu committed 5 years ago

b964bdb5

Fix workspace destruction ordering (#23096)

Yinghai Lu committed 5 years ago

2a8d5a13

Include module names in recursive error stacks (#22921)

davidriazati committed 5 years ago

79c4f83f

Quantization aware training in eager mode (#23082)

jerryzh168 committed 5 years ago

7cc029cb

Newer Older