DingXiaoH/RepLKNet-pytorch

Issues

• heatmap and erf

  #63 opened 3 months ago by cdzheng2g
  0

• ModuleNotFoundError: No module named '_depthwise_conv2d_implicit_gemm_C'

  #50 opened 2 years ago by XiaxiaBubble
  3

• Instructions about how to install 19_large_depthwise_conv2d_torch_extension

  #57 opened a year ago by liyiersan
  1

• Why ERF is using relu to cut off negative grad?

  #62 opened 7 months ago by hachreak
  0

• How does now run and pretrain

  #61 opened 9 months ago by ayushnangia
  0

• export to onnx failing with RuntimeError: invalid unordered_map<K, T> key

  #60 opened 10 months ago by santoshamd
  0

• No module named 'timm.optim.novograd‘

  #52 opened 2 years ago by guyunsu128
  3

• DepthWiseConv2dImplicitGEMM has no 'padding' class attribute(actually zero)

  #48 opened 2 years ago by zzzc18
  9

• ERF is smaller after changing the input image size

  #25 opened 3 years ago by sleepat-11
  6

• Using a transformer type model to calculate ERF yields very poor results

  #54 opened 2 years ago by 864530979
  0

• I try to increase the -1 dim ,the big kernal becomes more and more slower than conv2d

  #56 opened 2 years ago by Marsjunwang
  1

• Questions about re-param of models

  #55 opened 2 years ago by TravisL24
  0

• Welcome update to OpenMMLab 2.0

  #53 opened 2 years ago by vansin
  0

• Validation metric difference between saving state_dict() and the whole model.

  #51 opened 2 years ago by Grace625
  0

• Installing large_depthwise_conv2d_torch_extension failed on Windows 10

  #13 opened 3 years ago by zechengtang
  7

• Using DW 3x3 in stem block

  #49 opened 2 years ago by quannguyen268
  0

• About large depthwise conv2d kernel speed

  #47 opened 2 years ago by zzk2021
  3

• RuntimeError: Error compiling objects for extension

  #46 opened 2 years ago by hahejiang
  0

• failed when compiled cutlass-master

  #39 opened 3 years ago by Zhzhyd
  2

• How to extract intermediate features?

  #21 opened 3 years ago by KleinXin
  5

• depthwise_conv2d_implicit_gemm installation

  #45 opened 2 years ago by caikw0602
  0

• NotImplementedError HELP PLEASE!

  #44 opened 2 years ago by Crawlling
  0

• Two error during the compile for 19_large_depthwise_conv2d_torch_extension

  #34 opened 3 years ago by ewrfcas
  4

• Reproducibility for large kernel conv

  #43 opened 3 years ago by raihanarfi12-lab
  0

• Warning from setup.py

  #42 opened 3 years ago by Emily-core
  0

• windoes --Python setup.py install --user is fail

  #41 opened 3 years ago by duhongwen
  0

• CUDA out of memory when testing on cityspaces

  #40 opened 3 years ago by awfniewf
  0

• failed when compiled cutlass-master

  #38 opened 3 years ago by zhoujingyu13687306871
  0

• There is a big gap between the accuracy of Miou running in mmseg and that in the paper

  #35 opened 3 years ago by tthupup
  2

• How to use DepthWiseConv2dImplicitGEMM, I only have input channels, kernal, bais

  #28 opened 3 years ago by CccJ23
  4

• ERF small fix

  #31 opened 3 years ago by layumi
  1

• About seg fine-tuning

  #33 opened 3 years ago by heikeyuhuajia
  2

• Failed when installing 19_large_depthwise_conv2d_torch_extension

  #36 opened 3 years ago by Yightwing
  1

• About MobileNet shortcut

  #32 opened 3 years ago by kai3316
  0

• depthwise_conv2d_implicit_gemm slower than nn.Conv2d

  #17 opened 3 years ago by wdmwhh
  10

• Re-Parameterizing for switching

  #12 opened 3 years ago by BoltenWang-Meta
  1

• How to train on your own dataset

  #19 opened 3 years ago by BainOuO
  2

• setup.py build error

  #27 opened 3 years ago by carlsummer
  1

• KeyError: "CascadeRCNN: 'RepLKNet is not in the models registry

  #18 opened 3 years ago by Williamlizl
  1

• training for downstream task

  #22 opened 3 years ago by thancaocuong
  1

• Accuracy of EMA is even lower than the trained model

  #24 opened 3 years ago by Shiweiliuiiiiiii
  2

• how to make large kernel work?

  #29 opened 3 years ago by TheGreatTreatsby
  1

• bias

  #30 opened 3 years ago by hbw945
  1

• honestly, my computer CUDA is out of memory, maybe this model not suit for application.

  #26 opened 3 years ago by csliuchang
  1

• Get error when of using provided pytorch implementation of DWConv.

  #23 opened 3 years ago by jihaonew
  1

• Compiling 19_large_depthwise_conv2d_torch_extension

  #20 opened 3 years ago by gaowq2017
  3

• Is DepthWiseConv2dImplicitGEMM absolutely necessary?

  #16 opened 3 years ago by Harvey-Mei
  2

• Pre-trained weights of RepLKNet-13 for visualizing the ERF

  #15 opened 3 years ago by ShunLu91
  5

• the NET seems not working well with pretrained weight replknet31_base_224_pt1k_basecls.pkl

  #11 opened 3 years ago by dejavudejavu
  1

• IndexError: list index out of range

  #14 opened 3 years ago by Shiweiliuiiiiiii
  1