[Docs] 按照yolox_pose文档中的环境变量设置方式后进行模型转换存在错误 #2720

vicnoah · 2023-09-25T11:44:28Z

📚 The doc issue

执行命令

!export PYTHONPATH=/content/mmpose/projects/yolox_pose:$PYTHONPATH && echo $PYTHONPATH && cd /content/mmpose/projects/yolox_pose && python ../../../mmdeploy/tools/deploy.py \
../../../mmdeploy/configs/mmpose/pose-detection_yolox-pose_onnxruntime_dynamic.py \
configs/yolox-pose_m_4xb64-300e_coco.py \
../../../drive/MyDrive/yolox/yolox-pose_publish-cee405cd_20230925.pth ../../demo.jpg \
--work-dir ../../../drive/MyDrive/yolox/onnx --device cpu --dump-info

会出现错误提示

/content/mmpose/projects/yolox_pose:/env/python
Traceback (most recent call last):
  File "/content/mmpose/projects/yolox_pose/../../../mmdeploy/tools/deploy.py", line 335, in <module>
    main()
  File "/content/mmpose/projects/yolox_pose/../../../mmdeploy/tools/deploy.py", line 123, in main
    deploy_cfg, model_cfg = load_config(deploy_cfg_path, model_cfg_path)
  File "/usr/local/lib/python3.10/dist-packages/mmdeploy/utils/config_utils.py", line 29, in load_config
    configs = [_load_config(cfg) for cfg in args]
  File "/usr/local/lib/python3.10/dist-packages/mmdeploy/utils/config_utils.py", line 29, in <listcomp>
    configs = [_load_config(cfg) for cfg in args]
  File "/usr/local/lib/python3.10/dist-packages/mmdeploy/utils/config_utils.py", line 22, in _load_config
    cfg = mmengine.Config.fromfile(cfg)
  File "/usr/local/lib/python3.10/dist-packages/mmengine/config/config.py", line 461, in fromfile
    import_modules_from_strings(**cfg_dict['custom_imports'])
  File "/usr/local/lib/python3.10/dist-packages/mmengine/utils/misc.py", line 77, in import_modules_from_strings
    imported_tmp = import_module(imp)
  File "/usr/lib/python3.10/importlib/__init__.py", line 126, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "<frozen importlib._bootstrap>", line 1050, in _gcd_import
  File "<frozen importlib._bootstrap>", line 1027, in _find_and_load
  File "<frozen importlib._bootstrap>", line 1006, in _find_and_load_unlocked
  File "<frozen importlib._bootstrap>", line 688, in _load_unlocked
  File "<frozen importlib._bootstrap_external>", line 883, in exec_module
  File "<frozen importlib._bootstrap>", line 241, in _call_with_frames_removed
  File "/content/mmpose/projects/yolox_pose/models/__init__.py", line 5, in <module>
    from .yolox_pose_head import *  # noqa
  File "/content/mmpose/projects/yolox_pose/models/yolox_pose_head.py", line 13, in <module>
    from mmyolo.models.dense_heads import YOLOXHead, YOLOXHeadModule
  File "/usr/local/lib/python3.10/dist-packages/mmyolo/models/__init__.py", line 4, in <module>
    from .dense_heads import *  # noqa: F401,F403
  File "/usr/local/lib/python3.10/dist-packages/mmyolo/models/dense_heads/__init__.py", line 10, in <module>
    from .yolov7_head import YOLOv7Head, YOLOv7HeadModule, YOLOv7p6HeadModule
  File "/usr/local/lib/python3.10/dist-packages/mmyolo/models/dense_heads/yolov7_head.py", line 16, in <module>
    from ..task_modules.assigners.batch_yolov7_assigner import BatchYOLOv7Assigner
  File "/usr/local/lib/python3.10/dist-packages/mmyolo/models/task_modules/__init__.py", line 2, in <module>
    from .assigners import BatchATSSAssigner, BatchTaskAlignedAssigner
  File "/usr/local/lib/python3.10/dist-packages/mmyolo/models/task_modules/assigners/__init__.py", line 4, in <module>
    from .batch_task_aligned_assigner import BatchTaskAlignedAssigner
  File "/usr/local/lib/python3.10/dist-packages/mmyolo/models/task_modules/assigners/batch_task_aligned_assigner.py", line 9, in <module>
    from mmyolo.models.losses import bbox_overlaps
  File "/usr/local/lib/python3.10/dist-packages/mmyolo/models/losses/__init__.py", line 3, in <module>
    from .oks_loss import OksLoss
  File "/usr/local/lib/python3.10/dist-packages/mmyolo/models/losses/oks_loss.py", line 17, in <module>
    class OksLoss(nn.Module):
  File "/usr/local/lib/python3.10/dist-packages/mmengine/registry/registry.py", line 666, in _register
    self._register_module(module=module, module_name=name, force=force)
  File "/usr/local/lib/python3.10/dist-packages/mmengine/registry/registry.py", line 611, in _register_module
    raise KeyError(f'{name} is already registered in {self.name} '
KeyError: 'OksLoss is already registered in model at models.oks_loss'

Suggest a potential alternative/fix

No response

The text was updated successfully, but these errors were encountered:

Ben-Louis · 2023-09-26T02:33:08Z

Projects 中的 YOLOX-Pose 目前只与 mmyolo 0.5.0 兼容 (https://github.com/open-mmlab/mmpose/tree/dev-1.x/projects/yolox_pose#prerequisites). 请先安装 mmyolo 0.5.0 来使用该项目

vicnoah · 2023-09-26T03:09:50Z

Projects 中的 YOLOX-Pose 目前只与 mmyolo 0.5.0 兼容 (https://github.com/open-mmlab/mmpose/tree/dev-1.x/projects/yolox_pose#prerequisites). 请先安装 mmyolo 0.5.0 来使用该项目

好的，我试试

vicnoah · 2023-09-26T05:03:36Z

Projects 中的 YOLOX-Pose 目前只与 mmyolo 0.5.0 兼容 (https://github.com/open-mmlab/mmpose/tree/dev-1.x/projects/yolox_pose#prerequisites). 请先安装 mmyolo 0.5.0 来使用该项目

使用0.5.0又存在其它的报错了。

/content/mmpose/projects/yolox-pose:/env/python
09/26 05:02:02 - mmengine - WARNING - Failed to search registry with scope "mmpose" in the "Codebases" registry tree. As a workaround, the current "Codebases" registry in "mmdeploy" is used to build instance. This may cause unexpected failure when running the built modules. Please check whether "mmpose" is a correct scope, or whether the registry is initialized.
09/26 05:02:02 - mmengine - WARNING - Failed to search registry with scope "mmpose" in the "mmpose_tasks" registry tree. As a workaround, the current "mmpose_tasks" registry in "mmdeploy" is used to build instance. This may cause unexpected failure when running the built modules. Please check whether "mmpose" is a correct scope, or whether the registry is initialized.
09/26 05:02:02 - mmengine - WARNING - Failed to search registry with scope "mmyolo" in the "Codebases" registry tree. As a workaround, the current "Codebases" registry in "mmdeploy" is used to build instance. This may cause unexpected failure when running the built modules. Please check whether "mmyolo" is a correct scope, or whether the registry is initialized.
/usr/local/lib/python3.10/dist-packages/mmpose/utils/setup_env.py:79: UserWarning: The current default scope "mmyolo" is not "mmpose", `register_all_modules` will force the currentdefault scope to be "mmpose". If this is not expected, please set `init_default_scope=False`.
  warnings.warn('The current default scope '
Traceback (most recent call last):
  File "/usr/local/lib/python3.10/dist-packages/mmengine/config/config.py", line 106, in __getattr__
    value = super().__getattr__(name)
  File "/usr/local/lib/python3.10/dist-packages/addict/addict.py", line 67, in __getattr__
    return self.__getitem__(item)
  File "/usr/local/lib/python3.10/dist-packages/mmengine/config/config.py", line 135, in __getitem__
    return self.build_lazy(super().__getitem__(key))
  File "/usr/local/lib/python3.10/dist-packages/mmengine/config/config.py", line 102, in __missing__
    raise KeyError(name)
KeyError: 'codec'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/content/mmpose/projects/yolox-pose/../../../mmdeploy/tools/deploy.py", line 335, in <module>
    main()
  File "/content/mmpose/projects/yolox-pose/../../../mmdeploy/tools/deploy.py", line 129, in main
    export2SDK(
  File "/usr/local/lib/python3.10/dist-packages/mmdeploy/backend/sdk/export_info.py", line 353, in export2SDK
    pipeline_info = get_pipeline(deploy_cfg, model_cfg, work_dir, device)
  File "/usr/local/lib/python3.10/dist-packages/mmdeploy/backend/sdk/export_info.py", line 286, in get_pipeline
    preprocess = get_preprocess(deploy_cfg, model_cfg, device=device)
  File "/usr/local/lib/python3.10/dist-packages/mmdeploy/backend/sdk/export_info.py", line 197, in get_preprocess
    transforms = task_processor.get_preprocess()
  File "/usr/local/lib/python3.10/dist-packages/mmdeploy/codebase/mmpose/deploy/pose_detection.py", line 342, in get_preprocess
    model_cfg = process_model_config(self.model_cfg, [''], input_shape)
  File "/usr/local/lib/python3.10/dist-packages/mmdeploy/codebase/mmpose/deploy/pose_detection.py", line 39, in process_model_config
    codec = cfg.codec
  File "/usr/local/lib/python3.10/dist-packages/mmengine/config/config.py", line 1489, in __getattr__
    return getattr(self._cfg_dict, name)
  File "/usr/local/lib/python3.10/dist-packages/mmengine/config/config.py", line 110, in __getattr__
    raise AttributeError(f"'{self.__class__.__name__}' object has no "
AttributeError: 'ConfigDict' object has no attribute 'codec'

Ben-Louis · 2023-09-26T05:40:36Z

这个报错和 mmyolo 版本无关，应该是 mmdeploy 里面的实现有点问题。我们这两天看一下这个问题

Ben-Louis · 2023-09-27T11:16:54Z

请参考这个 PR 修改 mmdeploy 的代码。这个 PR 中的代码兼容 MMPose dev-1.x 分支中的 YOLOX-Pose

vicnoah · 2023-09-27T13:12:25Z

请参考这个 PR 修改 mmdeploy 的代码。这个 PR 中的代码兼容 MMPose dev-1.x 分支中的 YOLOX-Pose

好的，我试试

vicnoah · 2023-09-28T02:29:19Z

MPose dev-1.x 分支中的 YOLOX-Pose

我用dev-1.x分支的代码后还是存在此错误。

安装mmpose的方式:

git clone -b dev-1.x https://github.com/open-mmlab/mmpose.git
cd mmpose && pip install -r requirements.txt && pip install -v -e .

安装mmdeploy的方式:
1.安装mmdeploy

pip install mmdeploy==1.3.0
pip install mmdeploy-runtime==1.3.0
pip install onnxruntime==1.12.0
git clone -b v1.3.0 --recursive https://github.com/open-mmlab/mmdeploy.git

2.替换相关代码文件

mmdeploy/mmdeploy/codebase/mmpose/deploy/pose_detection.py
mmdeploy/mmdeploy/codebase/mmpose/deploy/pose_detection_model.py
mmdeploy/mmdeploy/codebase/mmpose/models/heads/yolox_pose_head.py

环境信息:

09/28 02:25:36 - mmengine - INFO - 

09/28 02:25:36 - mmengine - INFO - **********Environmental information**********
09/28 02:25:37 - mmengine - INFO - sys.platform: linux
09/28 02:25:37 - mmengine - INFO - Python: 3.10.12 (main, Jun 11 2023, 05:26:28) [GCC 11.4.0]
09/28 02:25:37 - mmengine - INFO - CUDA available: False
09/28 02:25:37 - mmengine - INFO - numpy_random_seed: 2147483648
09/28 02:25:37 - mmengine - INFO - GCC: x86_64-linux-gnu-gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
09/28 02:25:37 - mmengine - INFO - PyTorch: 2.0.1+cu118
09/28 02:25:37 - mmengine - INFO - PyTorch compiling details: PyTorch built with:
  - GCC 9.3
  - C++ Version: 201703
  - Intel(R) oneAPI Math Kernel Library Version 2022.2-Product Build 20220804 for Intel(R) 64 architecture applications
  - Intel(R) MKL-DNN v2.7.3 (Git Hash 6dbeffbae1f23cbbeae17adb7b5b13f1f37c080e)
  - OpenMP 201511 (a.k.a. OpenMP 4.5)
  - LAPACK is enabled (usually provided by MKL)
  - NNPACK is enabled
  - CPU capability usage: AVX2
  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.8, CUDNN_VERSION=8.7.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -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_DISABLE_GPU_ASSERTS=ON, TORCH_VERSION=2.0.1, 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=1, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, 

09/28 02:25:37 - mmengine - INFO - TorchVision: 0.15.2+cu118
09/28 02:25:37 - mmengine - INFO - OpenCV: 4.8.0
09/28 02:25:37 - mmengine - INFO - MMEngine: 0.8.3
09/28 02:25:37 - mmengine - INFO - MMCV: 2.0.1
09/28 02:25:37 - mmengine - INFO - MMCV Compiler: GCC 9.3
09/28 02:25:37 - mmengine - INFO - MMCV CUDA Compiler: 11.8
09/28 02:25:37 - mmengine - INFO - MMDeploy: 1.3.0+1132e82
09/28 02:25:37 - mmengine - INFO - 

09/28 02:25:37 - mmengine - INFO - **********Backend information**********
09/28 02:25:37 - mmengine - INFO - tensorrt:	None
09/28 02:25:37 - mmengine - INFO - ONNXRuntime:	1.12.0
09/28 02:25:37 - mmengine - INFO - ONNXRuntime-gpu:	None
09/28 02:25:37 - mmengine - INFO - ONNXRuntime custom ops:	Available
09/28 02:25:37 - mmengine - INFO - pplnn:	None
09/28 02:25:37 - mmengine - INFO - ncnn:	None
09/28 02:25:37 - mmengine - INFO - snpe:	None
09/28 02:25:37 - mmengine - INFO - openvino:	None
09/28 02:25:37 - mmengine - INFO - torchscript:	2.0.1+cu118
09/28 02:25:37 - mmengine - INFO - torchscript custom ops:	NotAvailable
09/28 02:25:37 - mmengine - INFO - rknn-toolkit:	None
09/28 02:25:37 - mmengine - INFO - rknn-toolkit2:	None
09/28 02:25:37 - mmengine - INFO - ascend:	None
09/28 02:25:37 - mmengine - INFO - coreml:	None
09/28 02:25:37 - mmengine - INFO - tvm:	None
09/28 02:25:37 - mmengine - INFO - vacc:	None
09/28 02:25:37 - mmengine - INFO - 

09/28 02:25:37 - mmengine - INFO - **********Codebase information**********
09/28 02:25:37 - mmengine - INFO - mmdet:	3.0.0rc6
09/28 02:25:37 - mmengine - INFO - mmseg:	None
09/28 02:25:37 - mmengine - INFO - mmpretrain:	None
09/28 02:25:37 - mmengine - INFO - mmocr:	None
09/28 02:25:37 - mmengine - INFO - mmagic:	None
09/28 02:25:37 - mmengine - INFO - mmdet3d:	None
09/28 02:25:37 - mmengine - INFO - mmpose:	1.1.0
09/28 02:25:37 - mmengine - INFO - mmrotate:	None
09/28 02:25:37 - mmengine - INFO - mmaction:	None
09/28 02:25:37 - mmengine - INFO - mmrazor:	None
09/28 02:25:37 - mmengine - INFO - mmyolo:	0.5.0

Ben-Louis · 2023-09-28T02:32:00Z

请问您部署 YOLOX-Pose 的时候使用的命令是哪条呢？我这边尝试

python tools/deploy.py configs/mmpose/pose-detection_yolox-pose_onnxruntime_dynamic.py ../mmpose/configs/body_2d_keypoint/yoloxpose/coco/yoloxpose_s_8xb32-300e_coco-640.py ../mmpose/ckpts/yoloxpose_s_8xb32-300e_coco-640-56c79c1f_20230829.pth ../mmpose/tests/data/coco/000000197388.jpg --work-dir logs/yoloxpose/s/onnx --test-img ../mmpose/tests/data/coco/000000197388.jpg --device cuda:0 --show --dump-info

是可以跑通部署的

vicnoah · 2023-09-28T02:39:22Z

请问您部署 YOLOX-Pose 的时候使用的命令是哪条呢？我这边尝试

python tools/deploy.py configs/mmpose/pose-detection_yolox-pose_onnxruntime_dynamic.py ../mmpose/configs/body_2d_keypoint/yoloxpose/coco/yoloxpose_s_8xb32-300e_coco-640.py ../mmpose/ckpts/yoloxpose_s_8xb32-300e_coco-640-56c79c1f_20230829.pth ../mmpose/tests/data/coco/000000197388.jpg --work-dir logs/yoloxpose/s/onnx --test-img ../mmpose/tests/data/coco/000000197388.jpg --device cuda:0 --show --dump-info

是可以跑通部署的

我是在谷歌colab环境下跑的，这是代码。

!export PYTHONPATH=/content/mmpose/projects/yolox_pose:$PYTHONPATH && echo $PYTHONPATH && cd /content/mmpose/projects/yolox_pose && python ../../../mmdeploy/tools/deploy.py \
../../../mmdeploy/configs/mmpose/pose-detection_yolox-pose_onnxruntime_dynamic.py \
configs/yolox-pose_m_4xb64-300e_coco.py \
../../../drive/MyDrive/yolox/yolox-pose_publish-cee405cd_20230925.pth ../../demo.jpg \
--work-dir ../../../drive/MyDrive/yolox/onnx --device cpu --dump-info

Ben-Louis · 2023-09-28T02:43:26Z

projects 里面的 YOLOX-Pose (mmpose/projects/yolox_pose) 后面应该不会继续维护了。建议使用主库中的 YOLOX-Pose，兼容性和性能都会更好一点。前面提到的 mmdeploy 的 PR 也是适配主库中的 YOLOX-Pose

vicnoah · 2023-09-28T03:02:44Z

projects 里面的 YOLOX-Pose (mmpose/projects/yolox_pose) 后面应该不会继续维护了。建议使用主库中的 YOLOX-Pose，兼容性和性能都会更好一点。前面提到的 mmdeploy 的 PR 也是适配主库中的 YOLOX-Pose

那不再需要设置PYTHONPATH了吗？我这个命令是正确的吗？新的代码弄进去还是在报错。

命令

!cd mmpose && python ../mmdeploy/tools/deploy.py \
../mmdeploy/configs/mmpose/pose-detection_yolox-pose_onnxruntime_dynamic.py \
configs/body_2d_keypoint/yoloxpose/coco/yoloxpose_tiny_4xb64-300e_coco-416.py \
../drive/MyDrive/yolox/yolox-pose_publish-cee405cd_20230925.pth demo.jpg \
--work-dir ../drive/MyDrive/yolox/onnx --device cpu --dump-info

Traceback (most recent call last):
  File "/usr/local/lib/python3.10/dist-packages/mmengine/config/config.py", line 107, in __getattr__
    value = super().__getattr__(name)
  File "/usr/local/lib/python3.10/dist-packages/addict/addict.py", line 67, in __getattr__
    return self.__getitem__(item)
  File "/usr/local/lib/python3.10/dist-packages/mmengine/config/config.py", line 136, in __getitem__
    return self.build_lazy(super().__getitem__(key))
  File "/usr/local/lib/python3.10/dist-packages/mmengine/config/config.py", line 103, in __missing__
    raise KeyError(name)
KeyError: 'bgr_to_rgb'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/content/mmpose/../mmdeploy/tools/deploy.py", line 335, in <module>
    main()
  File "/content/mmpose/../mmdeploy/tools/deploy.py", line 129, in main
    export2SDK(
  File "/usr/local/lib/python3.10/dist-packages/mmdeploy/backend/sdk/export_info.py", line 353, in export2SDK
    pipeline_info = get_pipeline(deploy_cfg, model_cfg, work_dir, device)
  File "/usr/local/lib/python3.10/dist-packages/mmdeploy/backend/sdk/export_info.py", line 286, in get_pipeline
    preprocess = get_preprocess(deploy_cfg, model_cfg, device=device)
  File "/usr/local/lib/python3.10/dist-packages/mmdeploy/backend/sdk/export_info.py", line 197, in get_preprocess
    transforms = task_processor.get_preprocess()
  File "/usr/local/lib/python3.10/dist-packages/mmdeploy/codebase/mmpose/deploy/pose_detection.py", line 342, in get_preprocess
    model_cfg = process_model_config(self.model_cfg, [''], input_shape)
  File "/usr/local/lib/python3.10/dist-packages/mmdeploy/codebase/mmpose/deploy/pose_detection.py", line 62, in process_model_config
    to_rgb=data_preprocessor.bgr_to_rgb))
  File "/usr/local/lib/python3.10/dist-packages/mmengine/config/config.py", line 111, in __getattr__
    raise AttributeError(f"'{self.__class__.__name__}' object has no "
AttributeError: 'ConfigDict' object has no attribute 'bgr_to_rgb'

Ben-Louis · 2023-09-28T07:06:40Z

根据报错信息

File "/usr/local/lib/python3.10/dist-packages/mmdeploy/codebase/mmpose/deploy/pose_detection.py", line 62, in process_model_config
to_rgb=data_preprocessor.bgr_to_rgb))

您现在使用的 mmdeploy 是用 pip 安装的预编译包，所以您修改的 mmdeploy 代码不会产生效果。您在改完 mmdeploy 的代码后需要执行一下 pip install -e /content/mmdeploy

vicnoah · 2023-09-28T11:27:09Z

根据报错信息
File "/usr/local/lib/python3.10/dist-packages/mmdeploy/codebase/mmpose/deploy/pose_detection.py", line 62, in process_model_config
to_rgb=data_preprocessor.bgr_to_rgb))
您现在使用的 mmdeploy 是用 pip 安装的预编译包，所以您修改的 mmdeploy 代码不会产生效果。您在改完 mmdeploy 的代码后需要执行一下 pip install -e /content/mmdeploy

非常感谢，问题已经解决了。

wenkaiH · 2023-10-16T08:55:16Z

mmdeploy/mmdeploy/codebase/mmpose/deploy/pose_detection.py
mmdeploy/mmdeploy/codebase/mmpose/deploy/pose_detection_model.py
mmdeploy/mmdeploy/codebase/mmpose/models/heads/yolox_pose_head.py

根据报错信息
File "/usr/local/lib/python3.10/dist-packages/mmdeploy/codebase/mmpose/deploy/pose_detection.py", line 62, in process_model_config
to_rgb=data_preprocessor.bgr_to_rgb))
您现在使用的 mmdeploy 是用 pip 安装的预编译包，所以您修改的 mmdeploy 代码不会产生效果。您在改完 mmdeploy 的代码后需要执行一下 pip install -e /content/mmdeploy

同样的问题，按上述办法更改mmdeploy的三个文件，重新pip install -e ，但执行命令后出现如下问题：

python tools/deploy.py configs/mmpose/pose-detection_yolox-pose_onnxruntime_dynamic.py ../mmpose/configs/body_2d_keypoint/yoloxpose/coco/yoloxpose_m_8xb32-300e_coco-640.py https://download.openmmlab.com/mmpose/v1/body_2d_keypoint/yolox_pose/yoloxpose_m_8xb32-300e_coco-640-84e9a538_20230829.pth demo/resources/human-pose.jpg --work-dir mmdeploy-model/yoloxpose-ort --device cpu --show --dump-info

10/16 16:49:28 - mmengine - WARNING - Failed to search registry with scope "mmpose" in the "Codebases" registry tree. As a workaround, the current "Codebases" registry in "mmdeploy" is used to build instance. This may cause unexpected failure when running the built modules. Please check whether "mmpose" is a correct scope, or whether the registry is initialized.
10/16 16:49:28 - mmengine - WARNING - Failed to search registry with scope "mmpose" in the "mmpose_tasks" registry tree. As a workaround, the current "mmpose_tasks" registry in "mmdeploy" is used to build instance. This may cause unexpected failure when running the built modules. Please check whether "mmpose" is a correct scope, or whether the registry is initialized.
10/16 16:49:30 - mmengine - INFO - Start pipeline mmdeploy.apis.pytorch2onnx.torch2onnx in subprocess
10/16 16:49:31 - mmengine - WARNING - Failed to search registry with scope "mmpose" in the "Codebases" registry tree. As a workaround, the current "Codebases" registry in "mmdeploy" is used to build instance. This may cause unexpected failure when running the built modules. Please check whether "mmpose" is a correct scope, or whether the registry is initialized.
10/16 16:49:31 - mmengine - WARNING - Failed to search registry with scope "mmpose" in the "mmpose_tasks" registry tree. As a workaround, the current "mmpose_tasks" registry in "mmdeploy" is used to build instance. This may cause unexpected failure when running the built modules. Please check whether "mmpose" is a correct scope, or whether the registry is initialized.
Process Process-2:
Traceback (most recent call last):
File "/home/ccnu-train/anaconda3/envs/mmdeploy/lib/python3.8/multiprocessing/process.py", line 315, in _bootstrap
self.run()
File "/home/ccnu-train/anaconda3/envs/mmdeploy/lib/python3.8/multiprocessing/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/ccnu-train/hwk/mmdeploy/mmdeploy/apis/core/pipeline_manager.py", line 107, in call
ret = func(*args, **kwargs)
File "/home/ccnu-train/hwk/mmdeploy/mmdeploy/apis/pytorch2onnx.py", line 63, in torch2onnx
torch_model = task_processor.build_pytorch_model(model_checkpoint)
File "/home/ccnu-train/hwk/mmdeploy/mmdeploy/codebase/base/task.py", line 120, in build_pytorch_model
model = MODELS.build(model)
File "/home/ccnu-train/anaconda3/envs/mmdeploy/lib/python3.8/site-packages/mmengine/registry/registry.py", line 570, in build
return self.build_func(cfg, *args, **kwargs, registry=self)
File "/home/ccnu-train/anaconda3/envs/mmdeploy/lib/python3.8/site-packages/mmengine/registry/build_functions.py", line 232, in build_model_from_cfg
return build_from_cfg(cfg, registry, default_args)
File "/home/ccnu-train/anaconda3/envs/mmdeploy/lib/python3.8/site-packages/mmengine/registry/build_functions.py", line 121, in build_from_cfg
obj = obj_cls(**args) # type: ignore
File "/home/ccnu-train/anaconda3/envs/mmdeploy/lib/python3.8/site-packages/mmpose/models/pose_estimators/bottomup.py", line 41, in init
super().init(
File "/home/ccnu-train/anaconda3/envs/mmdeploy/lib/python3.8/site-packages/mmpose/models/pose_estimators/base.py", line 43, in init
super().init(
File "/home/ccnu-train/anaconda3/envs/mmdeploy/lib/python3.8/site-packages/mmengine/model/base_model/base_model.py", line 78, in init
self.data_preprocessor = MODELS.build(data_preprocessor)
File "/home/ccnu-train/anaconda3/envs/mmdeploy/lib/python3.8/site-packages/mmengine/registry/registry.py", line 570, in build
return self.build_func(cfg, *args, **kwargs, registry=self)
File "/home/ccnu-train/anaconda3/envs/mmdeploy/lib/python3.8/site-packages/mmengine/registry/build_functions.py", line 232, in build_model_from_cfg
return build_from_cfg(cfg, registry, default_args)
File "/home/ccnu-train/anaconda3/envs/mmdeploy/lib/python3.8/site-packages/mmengine/registry/build_functions.py", line 121, in build_from_cfg
obj = obj_cls(**args) # type: ignore
TypeError: init() got an unexpected keyword argument 'batch_augments'
10/16 16:49:32 - mmengine - ERROR - /home/ccnu-train/hwk/mmdeploy/mmdeploy/apis/core/pipeline_manager.py - pop_mp_output - 80 - mmdeploy.apis.pytorch2onnx.torch2onnx with Call id: 0 failed. exit.

Ben-Louis · 2023-10-16T09:46:53Z

@wenkaiH，您好，请问您安装的 mm 系列库都是什么版本的呢？

wenkaiH · 2023-10-16T10:10:19Z

@wenkaiH，您好，请问您安装的 mm 系列库都是什么版本的呢？

mmcv 2.0.1 pypi_0 pypi
mmdeploy 1.3.0 dev_0
mmdeploy-runtime-gpu 1.3.0 pypi_0 pypi
mmdet 3.0.0 pypi_0 pypi
mmengine 0.8.4 pypi_0 pypi
mmpose 1.1.0 pypi_0 pypi
mmyolo 0.5.0 pypi_0 pypi

Ben-Louis · 2023-10-16T10:20:20Z

您安装的 mmpose 1.1.0 正式版里面还没有这个 YOLOX-Pose。可以试下在您的 mmpose 文件夹里执行一下 pip install -e . 或者将 mmpose 升级到 1.2.0

wenkaiH · 2023-10-16T10:31:56Z

您安装的 mmpose 1.1.0 正式版里面还没有这个 YOLOX-Pose。可以试下在您的 mmpose 文件夹里执行一下 pip install -e . 或者将 mmpose 升级到 1.2.0

尝试了pip install -e . 成功解决啦！非常感谢！！！

wenkaiH · 2023-10-16T11:30:40Z

您安装的 mmpose 1.1.0 正式版里面还没有这个 YOLOX-Pose。可以试下在您的 mmpose 文件夹里执行一下 pip install -e . 或者将 mmpose 升级到 1.2.0

接一个问题哈，部署成功后 API推理是没问题的，但 SDK推理会在加载pipeline报一些参数错：

loading libmmdeploy_trt_net.so ... failed to load library libmmdeploy_trt_net.so loading libmmdeploy_ort_net.so ... [2023-10-16 19:07:59.107] [mmdeploy] [info] [model.cpp:35] [DirectoryModel] Load model: "mmdeploy-model/yoloxpose-ort/" [2023-10-16 19:07:59.393] [mmdeploy] [error] [compose.cpp:37] Unable to find Transform creator: BottomupResize. Available transforms: [("CenterCrop", 0), ("Collect", 0), ("Compose", 0), ("DefaultFormatBundle", 0), ("FormatShape", 0), ("ImageToTensor", 0), ("LetterResize", 0), ("Lift", 0), ("LoadImageFromFile", 0), ("Normalize", 0), ("Pad", 0), ("RescaleToHeight", 0), ("Resize", 0), ("ResizeOCR", 0), ("ShortScaleAspectJitter", 0), ("TenCrop", 0), ("ThreeCrop", 0), ("TopDownAffine", 0), ("TopDownGetBboxCenterScale", 0)] [2023-10-16 19:07:59.393] [mmdeploy] [error] [task.cpp:99] error parsing config: { "context": { "device": "", "model": "", "stream": "" }, "input": [ "img" ], "module": "Transform", "name": "Preprocess", "output": [ "prep_output" ], "transforms": [ { "type": "LoadImageFromFile" }, { "input_size": [ 640, 640 ], "pad_val": [ 114, 114, 114 ], "type": "BottomupResize" }, { "mean": [ 0, 0, 0 ], "std": [ 1, 1, 1 ], "to_rgb": false, "type": "Normalize" }, { "keys": [ "img" ], "type": "ImageToTensor" }, { "keys": [ "img" ], "meta_keys": [ "img_shape", "pad_shape", "ori_shape", "img_norm_cfg", "scale_factor", "bbox_score", "center", "scale" ], "type": "Collect" } ], "type": "Task" } [2023-10-16 19:07:59.977] [mmdeploy] [error] [common.h:50] Could not found entry 'UNKNOWN' in mmpose. Available components: [("DeepposeRegressionHeadDecode", 0), ("SimCCLabelDecode", 0), ("TopdownHeatmapBaseHeadDecode", 0), ("TopdownHeatmapMSMUHeadDecode", 0), ("TopdownHeatmapMultiStageHeadDecode", 0), ("TopdownHeatmapSimpleHeadDecode", 0), ("ViPNASHeatmapSimpleHeadDecode", 0)] [2023-10-16 19:07:59.977] [mmdeploy] [error] [task.cpp:99] error parsing config: { "component": "UNKNOWN", "context": { "device": "", "model": "", "stream": "" }, "input": [ "prep_output", "infer_output" ], "module": "mmpose", "name": "postprocess", "output": [ "post_output" ], "params": { "flip_test": false, "input_size": [ 640, 640 ], "nms_thr": 0.65, "score_thr": 0.5, "type": "YOLOXPoseAnnotationProcessor" }, "type": "Task" } 段错误 (核心已转储)

vicnoah closed this as completed Sep 28, 2023

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[Docs] 按照yolox_pose文档中的环境变量设置方式后进行模型转换存在错误 #2720

[Docs] 按照yolox_pose文档中的环境变量设置方式后进行模型转换存在错误 #2720

vicnoah commented Sep 25, 2023

Ben-Louis commented Sep 26, 2023

vicnoah commented Sep 26, 2023

vicnoah commented Sep 26, 2023

Ben-Louis commented Sep 26, 2023

Ben-Louis commented Sep 27, 2023

vicnoah commented Sep 27, 2023

vicnoah commented Sep 28, 2023

Ben-Louis commented Sep 28, 2023 •

edited

Loading

vicnoah commented Sep 28, 2023

Ben-Louis commented Sep 28, 2023

vicnoah commented Sep 28, 2023 •

edited

Loading

Ben-Louis commented Sep 28, 2023 •

edited

Loading

vicnoah commented Sep 28, 2023

wenkaiH commented Oct 16, 2023

Ben-Louis commented Oct 16, 2023

wenkaiH commented Oct 16, 2023

Ben-Louis commented Oct 16, 2023

wenkaiH commented Oct 16, 2023

wenkaiH commented Oct 16, 2023

[Docs] 按照yolox_pose文档中的环境变量设置方式后进行模型转换存在错误 #2720

[Docs] 按照yolox_pose文档中的环境变量设置方式后进行模型转换存在错误 #2720

Comments

vicnoah commented Sep 25, 2023

📚 The doc issue

Suggest a potential alternative/fix

Ben-Louis commented Sep 26, 2023

vicnoah commented Sep 26, 2023

vicnoah commented Sep 26, 2023

Ben-Louis commented Sep 26, 2023

Ben-Louis commented Sep 27, 2023

vicnoah commented Sep 27, 2023

vicnoah commented Sep 28, 2023

Ben-Louis commented Sep 28, 2023 • edited Loading

vicnoah commented Sep 28, 2023

Ben-Louis commented Sep 28, 2023

vicnoah commented Sep 28, 2023 • edited Loading

Ben-Louis commented Sep 28, 2023 • edited Loading

vicnoah commented Sep 28, 2023

wenkaiH commented Oct 16, 2023

Ben-Louis commented Oct 16, 2023

wenkaiH commented Oct 16, 2023

Ben-Louis commented Oct 16, 2023

wenkaiH commented Oct 16, 2023

wenkaiH commented Oct 16, 2023

Ben-Louis commented Sep 28, 2023 •

edited

Loading

vicnoah commented Sep 28, 2023 •

edited

Loading

Ben-Louis commented Sep 28, 2023 •

edited

Loading