RuntimeError: Input type (torch.cuda.FloatTensor) and weight type (torch.FloatTensor) should be the same

[danilovabg]

Prerequisite

• I have searched Issues and Discussions but cannot get the expected help.
• I have read the FAQ documentation but cannot get the expected help.
• The bug has not been fixed in the latest version (master) or latest version (3.x).

Task

I have modified the scripts/configs, or I'm working on my own tasks/models/datasets.

Branch

master branch https://github.com/open-mmlab/mmdetection

Environment

sys.platform: linux
Python: 3.10.6 (main, Nov 14 2022, 16:10:14) [GCC 11.3.0]
CUDA available: True
GPU 0: NVIDIA GeForce RTX 3060
CUDA_HOME: /usr/local/cuda
NVCC: Cuda compilation tools, release 12.0, V12.0.140
GCC: x86_64-linux-gnu-gcc (Ubuntu 11.3.0-1ubuntu1~22.04) 11.3.0
PyTorch: 1.13.1+cu117
PyTorch compiling details: PyTorch built with:

• GCC 9.3
• C++ Version: 201402
• Intel(R) Math Kernel Library Version 2020.0.0 Product Build 20191122 for Intel(R) 64 architecture applications
• Intel(R) MKL-DNN v2.6.0 (Git Hash 52b5f107dd9cf10910aaa19cb47f3abf9b349815)
• OpenMP 201511 (a.k.a. OpenMP 4.5)
• LAPACK is enabled (usually provided by MKL)
• NNPACK is enabled
• CPU capability usage: AVX2
• CUDA Runtime 11.7
• 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_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
• CuDNN 8.5
• Magma 2.6.1
• Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.7, CUDNN_VERSION=8.5.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -fabi-version=11 -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 -Werror=non-virtual-dtor -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wunused-local-typedefs -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -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 -Werror=cast-function-type -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.13.1, USE_CUDA=ON, USE_CUDNN=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_ROCM=OFF,

TorchVision: 0.14.1+cu117
OpenCV: 4.7.0
MMCV: 1.7.1
MMCV Compiler: GCC 11.3
MMCV CUDA Compiler: not available
MMDetection: 2.28.1+

Reproduces the problem - code sample

RuntimeError Traceback (most recent call last)
Cell In[30], line 9
5 model = build_detector( cfg.model, train_cfg=cfg.get('train_cfg'), test_cfg=cfg.get('test_cfg'))
7 datasets = [build_dataset(cfg.data.train)]
----> 9 train_detector(model, datasets, cfg, distributed=False, validate=True)

File /usr/local/lib/python3.10/dist-packages/mmdet/apis/train.py:246, in train_detector(model, dataset, cfg, distributed, validate, timestamp, meta)
244 elif cfg.load_from:
245 runner.load_checkpoint(cfg.load_from)
--> 246 runner.run(data_loaders, cfg.workflow)

File ~/.local/lib/python3.10/site-packages/mmcv/runner/epoch_based_runner.py:136, in EpochBasedRunner.run(self, data_loaders, workflow, max_epochs, **kwargs)
134 if mode == 'train' and self.epoch >= self._max_epochs:
135 break
--> 136 epoch_runner(data_loaders[i], **kwargs)
138 time.sleep(1) # wait for some hooks like loggers to finish
139 self.call_hook('after_run')

File ~/.local/lib/python3.10/site-packages/mmcv/runner/epoch_based_runner.py:53, in EpochBasedRunner.train(self, data_loader, **kwargs)
51 self._inner_iter = i
52 self.call_hook('before_train_iter')
---> 53 self.run_iter(data_batch, train_mode=True, **kwargs)
54 self.call_hook('after_train_iter')
55 del self.data_batch
...
458 _pair(0), self.dilation, self.groups)
--> 459 return F.conv2d(input, weight, bias, self.stride,
460 self.padding, self.dilation, self.groups)

RuntimeError: Input type (torch.cuda.FloatTensor) and weight type (torch.FloatTensor) should be the same

Reproduces the problem - command or script

model = build_detector( cfg.model, train_cfg=cfg.get('train_cfg'), test_cfg=cfg.get('test_cfg'))
datasets = [build_dataset(cfg.data.train)]
train_detector(model, datasets, cfg, distributed=False, validate=True)

Reproduces the problem - error message

RuntimeError: Input type (torch.cuda.FloatTensor) and weight type (torch.FloatTensor) should be the same

Additional information

hello I use ssd300 model with modified backbone, dataset coco
at config I marked that device is cuda
cfg.gpu_ids = range(1)
cfg.device = 'cuda'

after that I use build_detector

model = build_detector( cfg.model, train_cfg=cfg.get('train_cfg'), test_cfg=cfg.get('test_cfg'))
datasets = [build_dataset(cfg.data.train)]
train_detector(model, datasets, cfg, distributed=False, validate=True)

at at start train get error
RuntimeError: Input type (torch.cuda.FloatTensor) and weight type (torch.FloatTensor) should be the same

[JamesYang568]

In fact, MMEngine can automatically change your module into cuda mode i.e. model = model.to(get_device()) in runner.py. Please double check your custom implementation like wrongly using list instead of modulelist for injection.