search

SwinTransformer / Swin-Transformer-Semantic-Segmentation

This is an official implementation for "Swin Transformer: Hierarchical Vision Transformer using Shifted Windows" on Semantic Segmentation.
https://arxiv.org/abs/2103.14030
Apache License 2.0
1.16k stars 223 forks source link

KeyError: "EncoderDecoder: 'SwinTransformer is not in the backbone registry'" #4

Closed YangParky closed 3 years ago

YangParky commented 3 years ago

When I am running the inference step, I meet the error as below.

微信图片_20210414193900

David-zaiwang commented 3 years ago

same problem， so how to solve it?

YangParky commented 3 years ago

not yet. I‘m still confused. But I manually added parameters to bypass this problem temporarily.

htzheng commented 3 years ago

Same problem here.

impiga commented 3 years ago

Hi, @Lature-Yang Could you please share your environment information for us to reproduce this bug? We tested on a machine with pytorch 1.6.0, cuda 10.1, mmcv 1.3.0 and everything worked well.

The following is our enviroment information:

sys.platform: linux
Python: 3.7.10 (default, Feb 26 2021, 18:47:35) [GCC 7.3.0]
CUDA available: True
GPU 0,1,2,3: Tesla V100-PCIE-16GB
CUDA_HOME: /usr/local/cuda
NVCC: Cuda compilation tools, release 10.1, V10.1.105
GCC: gcc (Ubuntu 5.4.0-6ubuntu1~16.04.12) 5.4.0 20160609
PyTorch: 1.6.0
PyTorch compiling details: PyTorch built with:
  - GCC 7.3
  - C++ Version: 201402
  - Intel(R) Math Kernel Library Version 2020.0.2 Product Build 20200624 for Intel(R) 64 architecture applications
  - Intel(R) MKL-DNN v1.5.0 (Git Hash e2ac1fac44c5078ca927cb9b90e1b3066a0b2ed0)
  - OpenMP 201511 (a.k.a. OpenMP 4.5)
  - NNPACK is enabled
  - CPU capability usage: AVX2
  - CUDA Runtime 10.1
  - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_37,code=compute_37
  - CuDNN 7.6.3
  - Magma 2.5.2
  - Build settings: BLAS=MKL, BUILD_TYPE=Release, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DUSE_VULKAN_WRAPPER -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-variable -Wno-unused-function -Wno-unused-result -Wno-unused-local-typedefs -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, USE_CUDA=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_STATIC_DISPATCH=OFF,

TorchVision: 0.7.0
OpenCV: 4.5.1
MMCV: 1.3.0
MMCV Compiler: GCC 7.3
MMCV CUDA Compiler: 10.1
MMSegmentation: 0.11.0+60ebe1a
Antonio-hi commented 3 years ago

Same issue

Antonio-hi commented 3 years ago 
cd Swin-Transformer-Semantic-Segmentation
pip install -e .  # or "python setup.py develop"

The main reason is that there is a custom package in this project. Compared with the original mmsegmentation, you need to import it manually. A simple alternative is to execute setup directly based on the project

YangParky commented 3 years ago 
cd Swin-Transformer-Semantic-Segmentation
pip install -e .  # or "python setup.py develop"

The main reason is that there is a custom package in this project. Compared with the original mmsegmentation, you need to import it manually. A simple alternative is to execute setup directly based on the project

Nice！！！Maybe I forgot this step.

YangParky commented 3 years ago

Hi, @Lature-Yang Could you please share your environment information for us to reproduce this bug? We tested on a machine with pytorch 1.6.0, cuda 10.1, mmcv 1.3.0 and everything worked well.

The following is our enviroment information:

sys.platform: linux
Python: 3.7.10 (default, Feb 26 2021, 18:47:35) [GCC 7.3.0]
CUDA available: True
GPU 0,1,2,3: Tesla V100-PCIE-16GB
CUDA_HOME: /usr/local/cuda
NVCC: Cuda compilation tools, release 10.1, V10.1.105
GCC: gcc (Ubuntu 5.4.0-6ubuntu1~16.04.12) 5.4.0 20160609
PyTorch: 1.6.0
PyTorch compiling details: PyTorch built with:
  - GCC 7.3
  - C++ Version: 201402
  - Intel(R) Math Kernel Library Version 2020.0.2 Product Build 20200624 for Intel(R) 64 architecture applications
  - Intel(R) MKL-DNN v1.5.0 (Git Hash e2ac1fac44c5078ca927cb9b90e1b3066a0b2ed0)
  - OpenMP 201511 (a.k.a. OpenMP 4.5)
  - NNPACK is enabled
  - CPU capability usage: AVX2
  - CUDA Runtime 10.1
  - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_37,code=compute_37
  - CuDNN 7.6.3
  - Magma 2.5.2
  - Build settings: BLAS=MKL, BUILD_TYPE=Release, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DUSE_VULKAN_WRAPPER -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-variable -Wno-unused-function -Wno-unused-result -Wno-unused-local-typedefs -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, USE_CUDA=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_STATIC_DISPATCH=OFF,

TorchVision: 0.7.0
OpenCV: 4.5.1
MMCV: 1.3.0
MMCV Compiler: GCC 7.3
MMCV CUDA Compiler: 10.1
MMSegmentation: 0.11.0+60ebe1a

The problem has been solved. Thanks ！！！