strisunshinewentingwang/pytorch

Pull Requests Commits

Use return names in JIT operators

cpuhrsch committed 5 years ago

b290a16b

Change ConvPoolOp<Context>::SetOutputSize to ConvPoolOp<Context>::GetOutputSize (#17764)

jerryzh168 committed 5 years ago

ac87488b

Clean up some old ScalarType stuff

Roy Li committed 5 years ago

cc7aec12

add reference to flake8-mypy in contributing.md

Elias Ellison committed 5 years ago

549eb9e9

Move lerp to ATen, add functionality for tensor weights (#17348)

vishwakftw committed 5 years ago

9d70e199

Refactor dispatcher (#17753)

izdeby committed 5 years ago

6227afb3

add layernorm to AD

wanchaol committed 5 years ago

aa57f178

move half<->float conversions to oss operators (#17548)

Hector Yuen committed 5 years ago

5bf9e419

Fix the update ONNX expect files (#17767)

houseroad committed 5 years ago

aa4c4c47

Cleanup testFusion/testOne: there are unused arguments.

Mikhail Zolotukhin committed 5 years ago

7bcc2301

Automatic update of fbcode/onnx to 96c58ceeacf0f2b73d752e413e4fd78787a12da3 (#17676)

houseroad committed 5 years ago

4480aa31

Set the default ONNX opset to the latest stable opset (i.e., 9) (#17736)

houseroad committed 5 years ago

1043ff6d

Add module attributes (#17309)

David Riazati committed 5 years ago

a2381fa3

- refactoring serialization of ONNX initializers to be name-based (#17420)

Spandan Tiwari committed 5 years ago

e4c9d750

ONNX Export for Max and Average Pooling in CEIL_MODE

lara-hdr committed 5 years ago

3f94fc48

use flake8-mypy (#17721)

Elias Ellison committed 5 years ago

561037ae

use fp16<->fp32 intrinsic (#17496)

jspark1105 committed 5 years ago

1d522598

Implement a Caffe2 standalone LSTM operator (#17726)

ahhegazy committed 5 years ago

f8778aef

caffe2:libtorch_cuda depends on caffe2:caffe2_gpu (#17729)

smessmer committed 5 years ago

7d02a1fb

add tensor and cost inference functions (#17684)

jspark1105 committed 5 years ago

39423fbd

ONNX Export Narrow op

lara-hdr committed 5 years ago

3dba1285

Keep the dim_type of hinted shape as BATCH if possible (#17734)

Yinghai Lu committed 5 years ago

32304046

fix different round behavior on CPU and GPU #16498 (#17443)

musikisomorphie committed 5 years ago

8ec73573

Warn about memory overlaps on expanded tensors (#17576)

zou3519 committed 5 years ago

68c5c668

fix exp fam. formula

ssnl committed 5 years ago

93768785

refactor caffe2 operator constructors - 10/9 (#17659)

smessmer committed 5 years ago

f6fda440

Improve ONNX symbolic for logsoftmax and softmax (#17672)

houseroad committed 5 years ago

4db3f8f8

Enable using CMD when building cpp extensions on Windows

peterjc123 committed 5 years ago

c78da0c6

Do not rename net boundary inputs/outputs during ssaRewrite. (#17545)

Yinghai Lu committed 5 years ago

a87d475c

Reapply D14078519 (#17596)

smessmer committed 5 years ago

9024faaa

Newer Older