strisunshinewentingwang/pytorch

Pull Requests Commits

Suppressing hypothesis health check for qnnpack_add

z-a-f committed 4 years ago

06a866de

Handle uninitialized min/max values in histogram observer (#27151)

raghuramank100 committed 4 years ago

9b22a554

Unify quantized conv and linear tests (#26992)

supriyar committed 4 years ago

f8d3eac4

Uninitialize the accumulation buffer to save some overhead (#27005)

jianyuh committed 4 years ago

68b4d22d

Fuse module enhancements (#26457)

raghuramank100 committed 4 years ago

66b73b09

Add control for observers in Fake-quantize module (#27113)

raghuramank100 committed 4 years ago

32eb3b8d

Support for add relu functional module (#26612)

raghuramank100 committed 4 years ago

5a2a34cd

Default observer and fake-quant for backends (#26627)

raghuramank100 committed 4 years ago

a8083e18

Emulate weight and activation only quant with fake quant, numerics test (#26625)

raghuramank100 committed 4 years ago

bc3fb36e

Quantization aware training: Freeze batch norm support (#26624)

raghuramank100 committed 4 years ago

e0822f10

Per channel fake quant (#26623)

raghuramank100 committed 4 years ago

cbfd4e05

Improve repr for quantized modules

James Reed committed 4 years ago

b9a2c8ac

Per-channel baseline (#26516)

raghuramank100 committed 4 years ago

6d7a73c0

Dont zero out buffers in dynamic linear (#27002)

James Reed committed 4 years ago

15e48276

fix AvgPool2d for 2^31-1 sized inputs, and get test_cuda_kernel_loop_overflow_large to working state

soumith committed 4 years ago

024fa347

fix segfault when printing error msg for list comp (#27398)

Krovatkin committed 4 years ago

95d2c7fc

Make `align_to` method-only. (#27304) (#27367)

zou3519 committed 4 years ago

7ba2baee

add AutoNonVariableTypeMode for USE_STATIC_DISPATCH on JIT->ATen path (#27274) (#27321)

ljk53 committed 4 years ago

ccf3a6de

Fixed Error message for tensor.align_to (#27221) (#27250)

izdeby committed 4 years ago

1ba6fc4c

Enabled comparison ops with named tensors (#27162) (#27249)

izdeby committed 4 years ago

d4d4bf56

Fix ONNX Interpolate (#27233)

houseroad committed 4 years ago

cee965fa

make class types callable (#26743) (#27226)

eellison committed 4 years ago

544c16cd

[jit] Fix `toIValue` dict iteration (#27112)

driazati committed 4 years ago

494a5563

Module method destroy (#27111)

IvanKobzarev committed 4 years ago

ba4c3a1c

Bump gloo (#27087)

pietern committed 4 years ago

c556f905

Turn on named tensor testing for v1.3.0 (#27084)

zou3519 committed 4 years ago

5c80dd3c

[jit] Serializing autograd ops into its own namespace (#27079)

wanchaol committed 4 years ago

2d8ee111

Serialize XLA Tensor (#27042)

ailzhang committed 4 years ago

ebc2519b

make cudnn rnn respect current stream (#27044)

ngimel committed 4 years ago

8c9e4b25

fix pytorch_linux_xenial_py3_6_gcc5_4_build for release branch

soumith committed 4 years ago

6a6f047f

Newer Older