Converting and testing onnx model

[nyanmn]

I am testing mmdeploy for converting model from pth to onnx.
Conversion is successful.
The following command is used.

python mmdeploy/tools/deploy.py \
    mmdeploy/configs/mmdet/detection/detection_onnxruntime_dynamic.py  \
    /workspace/mmaction2/demo/faster_rcnn_r50_fpn_2x_coco.py \
    /workspace/mmaction2/demo/faster_rcnn_r50_fpn_2x_coco_bbox_mAP-0.384_20200504_210434-a5d8aa15.pth \
    mmdetection/demo/demo.jpg \
    --work-dir /workspace/mmaction2/demo/onnx \
    --device cuda \
    --dump-info

deploy.json detail.json end2end.onnx pipeline.json files are produced from conversion.

When test with the following command

python mmdeploy/tools/test.py \
    mmdeploy/configs/mmdet/detection/detection_onnxruntime_dynamic.py \
    /workspace/mmaction2/demo/faster_rcnn_r50_fpn_2x_coco.py \
    --model /workspace/mmaction2/demo/onnx/end2end.onnx \
    --out /workspace/mmaction2/demo/out.pkl \
    --device cuda \
    --speed-test

I have error as

Traceback (most recent call last):
  File "mmdeploy/tools/test.py", line 151, in <module>
    main()
  File "mmdeploy/tools/test.py", line 108, in main
    dataset = task_processor.build_dataset(model_cfg, dataset_type)
  File "/usr/local/lib/python3.8/dist-packages/mmdeploy/codebase/base/task.py", line 93, in build_dataset
    dataset = self.codebase_class.build_dataset(dataset_cfg, dataset_type,
  File "/usr/local/lib/python3.8/dist-packages/mmdeploy/codebase/mmdet/deploy/mmdetection.py", line 78, in build_dataset
    dataset = build_dataset_mmdet(data_cfg)
  File "/workspace/mmdetection/mmdet/datasets/builder.py", line 82, in build_dataset
    dataset = build_from_cfg(cfg, DATASETS, default_args)
  File "/workspace/mmcv/mmcv/utils/registry.py", line 72, in build_from_cfg
    raise type(e)(f'{obj_cls.__name__}: {e}')
FileNotFoundError: CocoDataset: [Errno 2] No such file or directory: 'data/coco/annotations/instances_val2017.json'

When used different code as

from mmdeploy.apis import inference_model
result = inference_model(
  model_cfg='/workspace/mmaction2/demo/faster_rcnn_r50_fpn_2x_coco.py',
  deploy_cfg='/workspace/mmdeploy/configs/mmdet/detection/detection_onnxruntime_dynamic.py',
  backend_files=['/workspace/mmaction2/demo/onnx/end2end.onnx'],
  img='/workspace/mmdetection/demo/demo.jpg',
  device='cuda:0')

The different error came out as

Traceback (most recent call last):
  File "test_onnx.py", line 2, in <module>
    result = inference_model(
  File "/workspace/mmdeploy/mmdeploy/apis/inference.py", line 45, in inference_model
    model = task_processor.init_backend_model(backend_files)
  File "/workspace/mmdeploy/mmdeploy/codebase/mmdet/deploy/object_detection.py", line 73, in init_backend_model
    model = build_object_detection_model(
  File "/workspace/mmdeploy/mmdeploy/codebase/mmdet/deploy/object_detection_model.py", line 790, in build_object_detection_model
    backend_detector = __BACKEND_MODEL.build(
  File "/workspace/mmcv/mmcv/utils/registry.py", line 237, in build
    return self.build_func(*args, **kwargs, registry=self)
  File "/workspace/mmdeploy/mmdeploy/codebase/mmdet/deploy/object_detection_model.py", line 27, in __build_backend_model
    return registry.module_dict[partition_name](
  File "/workspace/mmdeploy/mmdeploy/codebase/mmdet/deploy/object_detection_model.py", line 64, in __init__
    self._init_wrapper(
  File "/workspace/mmdeploy/mmdeploy/codebase/mmdet/deploy/object_detection_model.py", line 78, in _init_wrapper
    self.wrapper = BaseBackendModel._build_wrapper(
  File "/workspace/mmdeploy/mmdeploy/codebase/base/backend_model.py", line 59, in _build_wrapper
    from mmdeploy.backend.onnxruntime import ORTWrapper
ImportError: cannot import name 'ORTWrapper' from 'mmdeploy.backend.onnxruntime' (/workspace/mmdeploy/mmdeploy/backend/onnxruntime/__init__.py)

How can I test onnx model?