Use of Focal Loss: RuntimeError sigmoid_focal_loss_forward_impl: implementation for device cuda:0 not found.

[Zihang-Wei]

Hello All, how should I use the focal loss inside the config file? It pops up this strange message for me... In the config file:

model = dict(
    test_cfg=dict(crop_size=(128, 128), stride=(85, 85)), 
    auxiliary_head=dict(
        num_classes=3,
        loss_decode=[
            # dict(type='CrossEntropyLoss', loss_name='loss_ce'),
            # dict(type='DiceLoss', loss_name='loss_dice', loss_weight=3.0),
            dict(type='FocalLoss', loss_name='loss_focal', loss_weight=3.0, gamma=2.0, alpha=[0.2, 0.1, 0.7])
            ]),
    decode_head=dict(
        num_classes=3, 
        loss_decode=[
            # dict(type='CrossEntropyLoss', loss_name='loss_ce', loss_weight=1.0),
            # dict(type='DiceLoss', loss_name='loss_dice', loss_weight=3.0),
            dict(type='FocalLoss', loss_name='loss_focal', loss_weight=3.0, gamma=2.0, alpha=[0.2, 0.1, 0.7]),
            # dict(type='LovaszLoss', loss_name='loss_lovasz', loss_weight=3.0)
            ]))

@RockeyCoss @Junjun2016 @MengzhangLI @xvjiarui @kahkeng

Originally posted by @Zihang-Wei in https://github.com/open-mmlab/mmsegmentation/issues/1024#issuecomment-1298185168

[Zihang-Wei]

The problem is solved by altering the DockerFile. Since there the mmcv is installed instead of mmcv-full althogh it is intended to install mmcv-full.

Just add before installing mmcv-full.

I suggest you to change the DockerFile. @xiexinch

[xiexinch]

Hi @Zihang-Wei, Thanks for your bug report, we'll fix it soon. Could you provide more information about your running environment? Since we would like to do more test.

[Zihang-Wei]

Sure~ Now after adding these two lines it is:

sys.platform: linux
Python: 3.8.12 (default, Oct 12 2021, 13:49:34) [GCC 7.5.0]
CUDA available: True
GPU 0: NVIDIA GeForce GTX 1080 Ti
CUDA_HOME: /usr/local/cuda
NVCC: Cuda compilation tools, release 11.3, V11.3.109
GCC: gcc (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0
PyTorch: 1.11.0
PyTorch compiling details: PyTorch built with:
  - GCC 7.3
  - C++ Version: 201402
  - Intel(R) oneAPI Math Kernel Library Version 2021.4-Product Build 20210904 for Intel(R) 64 architecture applications
  - Intel(R) MKL-DNN v2.5.2 (Git Hash a9302535553c73243c632ad3c4c80beec3d19a1e)
  - OpenMP 201511 (a.k.a. OpenMP 4.5)
  - LAPACK is enabled (usually provided by MKL)
  - NNPACK is enabled
  - CPU capability usage: AVX2
  - CUDA Runtime 11.3
  - 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_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;-gencode;arch=compute_37,code=compute_37
  - CuDNN 8.2
  - Magma 2.5.2
  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.3, CUDNN_VERSION=8.2.0, CXX_COMPILER=/opt/rh/devtoolset-7/root/usr/bin/c++, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -DEDGE_PROFILER_USE_KINETO -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-function -Wno-unused-result -Wno-unused-local-typedefs -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -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, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.11.0, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=OFF, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, 

TorchVision: 0.12.0
OpenCV: 4.6.0
MMCV: 1.6.2
MMCV Compiler: GCC 7.5
MMCV CUDA Compiler: 11.3
MMSegmentation: 0.29.0+

Before the MMCV CUDA Compiler is None

[Sere1nz]

I met the same problem. What should I do if Im not using docker but only on linux? I tried installing mmcv-full,but then when I install mmsegmentation, it said "AssertionError: MMCV==1.7.1 is used but incompatible. Please install mmcv>=2.0.0rc4."

But If i use mmcv==2.1.0 without "full", this "Use of Focal Loss: RuntimeError sigmoid_focal_loss_forward_impl: implementation for device cuda:0 not found." pops up. What should I do?