Cat example does not work out of the box

[daninaga]

Prerequisite

• I have searched the existing and past issues 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.

🐞 Describe the bug

I am able to consistently run the demos through:

from mmdet.apis import DetInferencer

# Initialize the DetInferencer
inferencer = DetInferencer('rtmdet_tiny_8xb32-300e_coco')
print(inferencer)

# Perform inference
inferencer('demo/demo.jpg', show=True)`

However whenever I try running the fine-tuning example for MMYOLO that uses YOLOv5 i always run into:

python tools/train.py configs/yolov5/yolov5_s-v61_fast_1xb12-40e_cat.py

...

miniconda3\envs\mmyolo\lib\site-packages\albumentations\core\composition.py", line 316, in _check_args
    raise ValueError(msg)
ValueError: Key img_path is not in available keys.

is there any need to modify the cat dataset provided or maybe there is some error in the config of yolov5_s-v61_fast_1xb12-40e_cat.py ?

Environment

sys.platform: win32
Python: 3.8.19 (default, Mar 20 2024, 19:55:45) [MSC v.1916 64 bit (AMD64)]
CUDA available: True
MUSA available: False
numpy_random_seed: 2147483648
GPU 0: NVIDIA GeForce RTX 3060
CUDA_HOME: C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.7
NVCC: Cuda compilation tools, release 11.7, V11.7.64
MSVC: Microsoft (R) C/C++ Optimizing Compiler Version 19.32.31328 for x64
GCC: n/a
PyTorch: 2.0.0
PyTorch compiling details: PyTorch built with:

• C++ Version: 199711
• MSVC 193431937
• Intel(R) Math Kernel Library Version 2020.0.2 Product Build 20200624 for Intel(R) 64 architecture applications
• Intel(R) MKL-DNN v2.7.3 (Git Hash 6dbeffbae1f23cbbeae17adb7b5b13f1f37c080e)
• OpenMP 2019
• LAPACK is enabled (usually provided by MKL)
• 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_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 8.5
• Magma 2.5.4
• Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.7, CUDNN_VERSION=8.5.0, CXX_COMPILER=C:/cb/pytorch_1000000000000/work/tmp_bin/sccache-cl.exe, CXX_FLAGS=/DWIN32 /D_WINDOWS /GR /EHsc /w /bigobj /FS -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOCUPTI -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_DISABLE_GPU_ASSERTS=OFF, TORCH_VERSION=2.0.0, 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=OFF, USE_NNPACK=OFF, USE_OPENMP=ON, USE_ROCM=OFF,

TorchVision: 0.15.0
OpenCV: 4.9.0
MMEngine: 0.10.4
MMCV: 2.0.1
MMDetection: 3.3.0
MMYOLO: 0.6.0+8c4d9dc

Additional information

No response

[gogoworker]

I'm having the same issue, do you have any solution to it, please

[daninaga]

apparently it is an issue with the MMYOLO backend when passing parameters to the albuminations.
workaround is to disable check in albumination core (C:\Users\{USER_NAME}\miniconda3\envs\{ENV_NAME}\Lib\site-packages\albumentations\core\composition.py):

    def _check_args(self, **kwargs: Any) -> None:
        checked_single = ["image", "mask"]
        checked_multi = ["masks"]
        check_bbox_param = ["bboxes"]
        check_keypoints_param = ["keypoints"]
        shapes = []
        for data_name, data in kwargs.items():
            # if data_name=='bboxes':
            #     print(data_name)
            #     print(data)
            if data_name not in self._available_keys and data_name not in ["mask", "masks"] and False:
                msg = f"Key {data_name} is not in available keys."
                # print('keys')
                # print(self._available_keys)
                # print(data)
                # print(kwargs.items())
                raise ValueError(msg)

[gogoworker]

Helo, I commented directly and simply and crudely on “raise ValueError(msg)”, and then “continue” solved the problem, I hope it helps you

[wdzwdxy]

补充：
禁用代码片段

def _check_args(self, **kwargs: Any) -> None:
        checked_single = ["image", "mask"]
        checked_multi = ["masks"]
        check_bbox_param = ["bboxes"]
        check_keypoints_param = ["keypoints"]
        shapes = []
        for data_name, data in kwargs.items():
            # if data_name not in self._available_keys and data_name not in ["mask", "masks"]:
            #     msg = f"Key {data_name} is not in available keys."
            #     raise ValueError(msg)
            internal_data_name = self._additional_targets.get(data_name, data_name)