strisunshinewentingwang/pytorch

Pull Requests Commits

refactor caffe2 operator constructors - 11/9 (#17722)

smessmer committed 5 years ago

4a7c549e

Suppress C408 lint (don't use dict constructor) (#17813)

ezyang committed 5 years ago

55cf9c74

Add matches_jit_signature to recent native functions

cpuhrsch committed 5 years ago

11f50e73

Add /MD to prevent linking errors on Windows

peterjc123 committed 5 years ago

fe90ee9d

Change message on unknown db type to be friendly (#17795)

Dmytro Dzhulgakov committed 5 years ago

a60fadfb

Trace rnn max_batch_size (#17727)

David Riazati committed 5 years ago

667763a6

Remove legacy way of exposing caffe2 operators to PyTorch (#17742)

smessmer committed 5 years ago

7f7d1285

Remove 'Tensor' key from ATen codegen. (#17782)

gchanan committed 5 years ago

b132f0f1

Remove ProcessorSpecificPlugin. (#17789)

gchanan committed 5 years ago

5ffd7dbb

Remove THPPlugin. (#17790)

gchanan committed 5 years ago

33c83a3f

Replace tens with hundreds.

ezyang committed 5 years ago

424a0318

Support failback for more operators in ideep (#17747)

Tim Khatkevich committed 5 years ago

6aacc1b2

Cleanup include files in jit/passes/common_subexpression_elimination.h.

Mikhail Zolotukhin committed 5 years ago

25692352

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

Newer Older