Issues
- 1
the device used in training
#55 opened by tianlianghai - 0
FasterNet 在GPU, CPU, and ARM processors上的速度是直接基于pytorch环境,还是使用了Tensorrt、OpenVINO、ONNX进行了部署
#64 opened by wsy-yjys - 0
这样的配置文件怎么在pcconv里融合注意力机制呢
#63 opened by Weeknd0214 - 3
怎样采用Pconv+Pwconv替换常规的3x3卷积?
#51 opened by xiaoche-24 - 1
Configuration steps for other datasets
#61 opened by mojeedoyedeji - 10
About the use of pconv in resnet
#34 opened by okt7760 - 0
Partial_conv3和group convolution
#62 opened by 23jisuper - 0
onnx转为trt后,速度提升不明显
#58 opened by xiaoche-24 - 0
为什么要把pooling层放在1x1卷积之前?
#60 opened by puyiwen - 0
- 2
Export to ONNX
#16 opened by WailordHe - 0
更改fasternet output size对性能的影响
#57 opened by DrunkedPiano - 0
- 0
- 0
Why are you executing this norm_layer = getattr(self, f'norm{idx}') x_out = norm_layer(x)?
#53 opened by nobody2231 - 0
pconv+pwconv
#52 opened by YuchaoXu - 1
replace the convolution in Resnet with PConv
#41 opened by wanderer1230 - 0
I can't find the Appendix of this paper!
#50 opened by haiduo - 1
How to get a pretrained weight
#49 opened by hefangnan - 0
- 7
Multi GPU training
#18 opened by VV1314 - 4
Some problems in section 3.1
#42 opened by juan19941228 - 8
About the experiment in the paper
#30 opened by 870231652 - 4
Question about the formula
#26 opened by 870231652 - 1
Completely replace Resnet18 with FasterNet?
#47 opened by Embarrassing1 - 2
about load model
#44 opened by 1072010918 - 1
problem with dataset
#43 opened by 1072010918 - 1
a question
#39 opened by nince9 - 0
- 0
Can it be accelerated with Tensorrt?
#45 opened by zzlnys - 6
Hello author, I have some questions about the ablation experiment section of your paper.
#27 opened by csjhhuc - 1
- 12
question about formula 7
#32 opened by abcsimple - 6
slicing slower than split_cat
#36 opened by wsy-yjys - 5
Some questions about “PConv + PWConv“
#31 opened by LKAMING97 - 1
Readme file improvement
#37 opened by yangshaobo0634 - 2
Can you provide the code for latency test?
#38 opened by leexinhao - 3
some question
#35 opened by jackiezhang11 - 1
- 1
The order of the blocks in the Stage
#29 opened by wsy-yjys - 2
A large pre-training weight
#28 opened by zhu011 - 7
Limitation to Partial_conv3
#21 opened by liamsun2019 - 1
About PWConv
#25 opened by xyh001007 - 1
Can it be used on mmdetection?
#24 opened by okt7760 - 1
- 5
why the model size all so big?
#22 opened by lucasjinreal - 2
Is it okay to increase the hierarchical stages?
#20 opened by bboogi - 1
some question about reparameterization in pconv
#19 opened by 870231652 - 1
PConv is similar to GhostConv
#17 opened by baowenshuwh - 2
How to measure the FLOPS the the model ?
#15 opened by jimmylin0979