pytorch QNNPACK issues - Githubissues

pytorch / QNNPACK

Quantized Neural Network PACKage - mobile-optimized implementation of quantized neural network operators

https://code.fb.com/ml-applications/qnnpack/

Other

1.52k stars 219 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Qnnpack accuracy very poor on unet model

#73 Amitdedhia6 opened 4 years ago
3
Support for nn.ConvTranspose2d

#72 Amitdedhia6 opened 4 years ago
0
Build failed on rasp pi 4 (Error: selected processor does not support `yield' in ARM mode)

#71 pyjhzwh closed 4 years ago
1
Understanding XZP ukernel

#70 wenfahu opened 4 years ago
1
Is QNNPACK still being developed?

#69 gibiansky opened 4 years ago
1
How to build dependencies separately

#68 adamjstewart opened 4 years ago
0
COMPAIR ERROR

#67 mathpopo closed 4 years ago
2
Users/ashkan/requantize

#66 AshkanAliabadi closed 5 years ago
3
Fix clamp operator to work when output_min == output_max.

#65 supriyar closed 5 years ago
0
Prepare to build on Windows

#64 dreiss closed 5 years ago
1
Add support for batch size 0

#63 dreiss closed 5 years ago
1
is there any example to use fp16 gemm？

#62 HaoLiuHust opened 5 years ago
0
build error

#61 lucasjinreal closed 4 years ago
0
Is the code multithreaded for arm64?

#60 RuABraun closed 5 years ago
7
3d convolution support

#59 dmenig closed 5 years ago
1
Prepare Qnnpack to build on Windows with MSVS

#58 boguscoder closed 5 years ago
0
Remove IGNORE_CODE_ALIGN_DIRECTIVES from aarch32 code

#57 alexbdv closed 5 years ago
0
Add IGNORE_CODE_ALIGN_DIRECTIVES macro that disables code align diretives

#56 alexbdv closed 5 years ago
0
Typo in the announcement blog?

#55 zhenhuaw-me closed 5 years ago
1
Q8GEMM per-channel quant 32bit/16bit accumulation

#54 mortzur opened 5 years ago
2
Crashed when running conv3x3 test

#53 wonderzy closed 5 years ago
0
Request to add a Concat layer

#52 E1eMenta closed 5 years ago
1
Per channel quant

#51 mortzur opened 5 years ago
0
What's the relationship between the ASM and C version micro kernel?

#50 zhenhuaw-me closed 5 years ago
5
QNNPACK | Prepare Qnnpack to build on Windows

#49 boguscoder closed 5 years ago
1
Request to add Conv1d related operation

#48 viggi1000 closed 5 years ago
1
No performance difference between ARMv7 (32-bit) Android and ARM64 (64-bit) Android

#47 yangwenca closed 5 years ago
1
Deconvolutional Layer is slow for large kernel size

#46 yangwenca closed 5 years ago
2
Implementation detail things...

#45 zhenhuaw-me closed 5 years ago
9
Per channel quantization for the weights

#44 RuABraun closed 5 years ago
1
Will you going to open source QNNPACK quantize tool and methods?

#43 csuhawk closed 5 years ago
2
No documentation? or how to execute a layer in QNNPACK

#42 RuABraun closed 5 years ago
8
Interested in using QNNPACK operators as part of another C++ library

#41 RuABraun closed 5 years ago
2
Enable xzp ukernels on Cortex-A76

#40 hlu1 closed 5 years ago
0
minor sp: directorory --> directory

#39 brettkoonce closed 5 years ago
1
Can you provide resnet_50_quant example?

#38 mosjoker closed 5 years ago
1
Semantic segmentation examples

#37 brucechou1983 closed 5 years ago
1
multi-thread problem.

#36 32l closed 5 years ago
7
How can I create a quantized model by QNNPACK?

#35 Meicsu199345 closed 5 years ago
2
how can I use qnnpack to recognize a cat?

#34 dvlee1024 closed 5 years ago
2
8bit depthwise convolution is much slow than tf-lite 8bit!!!

#33 32l closed 5 years ago
2
Could you pls provide Android Apk demo?

#32 Meicsu199345 closed 5 years ago
6
caffe2 abort when running benchmark with input_type uint8_t

#31 zhenhuaw-me closed 5 years ago
4
readme: update android benchmark commands

#30 zhenhuaw-me closed 5 years ago
1
Weird performance gap observed for Jetson TX2

#29 TerryTsao closed 5 years ago
7
Out-of-memory during compiling on Raspbian

#28 sheh closed 5 years ago
4
5x5 sse2 kernel and new interface support

#27 harouwu closed 5 years ago
1
5x5 convolution fix and make stack allocation

#26 harouwu closed 5 years ago
0
fix non1 batch size 5x5 buffer allocation

#25 harouwu closed 5 years ago
0
PEP bench in README

#24 harouwu closed 5 years ago
0