API Usage Error (Parameter check failed

[MyraBaba]

Env

• GPU, 2080RTX Ti
• docker nvcr.io/nvidia/tensorrt:22.07-py3
• 11.2
• TensorRT 8.4.1.

About this repo

yolov5

Your problem

root@23fd790e8b4b:/tensorrtx/yolov5/build# ./yolov5 -s yolov5s.wts yolov5s.engine s [08/20/2022-13:19:03] [W] [TRT] The implicit batch dimension mode has been deprecated. Please create the network with NetworkDefinitionCreationFlag::kEXPLICIT_BATCH flag whenever possible. Loading weights: yolov5s.wts [08/20/2022-13:19:03] [E] [TRT] 3: [network.cpp::addScale::736] Error Code 3: API Usage Error (Parameter check failed at: optimizer/api/network.cpp::addScale::736, condition: shift.count > 0 ? (shift.values != nullptr) : (shift.values == nullptr) ) yolov5: /tensorrtx/yolov5/common.hpp:155: nvinfer1::IScaleLayer* addBatchNorm2d(nvinfer1::INetworkDefinition*, std::map<std::__cxx11::basic_string, nvinfer1::Weights>&, nvinfer1::ITensor&, std::string, float): Assertion `scale_1' failed. Aborted (core dumped)

[wang-xinyu]

Did you try the official trained .pt model? https://github.com/wang-xinyu/tensorrtx/tree/master/yolov5#different-versions-of-yolov5

[MyraBaba]

@wang-xinyu Hi,

I followed the tutorial and create yolov5s.wts from official model. than follow instruction. here the error:

./yolov5 -s yolov5s.wts yolov5s.engine s [08/22/2022-11:49:11] [W] [TRT] The implicit batch dimension mode has been deprecated. Please create the network with NetworkDefinitionCreationFlag::kEXPLICIT_BATCH flag whenever possible. Loading weights: yolov5s.wts CUDA error 222 at /tensorrtx/yolov5/yololayer.cu:37yolov5: /tensorrtx/yolov5/yololayer.cu:37: nvinfer1::YoloLayerPlugin::YoloLayerPlugin(int, int, int, int, const std::vectorYolo::YoloKernel&): Assertion `0' failed. Aborted (core dumped)

docker nvcr.io/nvidia/tensorrt:22.07-py3 & TensorRT 8.4.1.

[MyraBaba]

@wang-xinyu

I changed to the tensorRT 7

than above model conversion and detections worked.

But yolo5m and yolo5l has error.

./yolov5ORG -s yolov5m-v4.0/yolov5m.wts yolov5m.engine s Loading weights: yolov5m-v4.0/yolov5m.wts [08/22/2022-23:42:07] [E] [TRT] Parameter check failed at: ../builder/Network.cpp::addScale::494, condition: shift.count > 0 ? (shift.values != nullptr) : (shift.values == nullptr) yolov5ORG: /tensorrtx/yolov5/common.hpp:155: nvinfer1::IScaleLayer* addBatchNorm2d(nvinfer1::INetworkDefinition*, std::map<std::__cxx11::basic_string, nvinfer1::Weights>&, nvinfer1::ITensor&, std::string, float): Assertion `scale_1' failed. Aborted (core dumped)

And for yolo5l

./yolov5ORG -s yolov5l.wts yolov5l.engine s Loading weights: yolov5l.wts [08/22/2022-23:42:53] [E] [TRT] (Unnamed Layer* 0) [Convolution]: kernel weights has count 6912 but 3456 was expected [08/22/2022-23:42:53] [E] [TRT] (Unnamed Layer* 0) [Convolution]: count of 6912 weights in kernel, but kernel dimensions (6,6) with 3 input channels, 32 output channels and 1 groups were specified. Expected Weights count is 3 * 66 * 32 / 1 = 3456 [08/22/2022-23:42:53] [E] [TRT] (Unnamed Layer 0) [Convolution]: kernel weights has count 6912 but 3456 was expected [08/22/2022-23:42:53] [E] [TRT] (Unnamed Layer* 0) [Convolution]: count of 6912 weights in kernel, but kernel dimensions (6,6) with 3 input channels, 32 output channels and 1 groups were specified. Expected Weights count is 3 * 66 * 32 / 1 = 3456 [08/22/2022-23:42:53] [E] [TRT] (Unnamed Layer 0) [Convolution]: kernel weights has count 6912 but 3456 was expected [08/22/2022-23:42:53] [E] [TRT] (Unnamed Layer* 0) [Convolution]: count of 6912 weights in kernel, but kernel dimensions (6,6) with 3 input channels, 32 output channels and 1 groups were specified. Expected Weights count is 3 * 66 * 32 / 1 = 3456 [08/22/2022-23:42:53] [E] [TRT] (Unnamed Layer 0) [Convolution]: kernel weights has count 6912 but 3456 was expected [08/22/2022-23:42:53] [E] [TRT] (Unnamed Layer* 0) [Convolution]: count of 6912 weights in kernel, but kernel dimensions (6,6) with 3 input channels, 32 output channels and 1 groups were specified. Expected Weights count is 3 * 66 * 32 / 1 = 3456 [08/22/2022-23:42:53] [E] [TRT] (Unnamed Layer 0) [Convolution]: kernel weights has count 6912 but 3456 was expected [08/22/2022-23:42:53] [E] [TRT] (Unnamed Layer* 0) [Convolution]: count of 6912 weights in kernel, but kernel dimensions (6,6) with 3 input channels, 32 output channels and 1 groups were specified. Expected Weights count is 3 * 66 * 32 / 1 = 3456 [08/22/2022-23:42:53] [E] [TRT] (Unnamed Layer 0) [Convolution]: kernel weights has count 6912 but 3456 was expected [08/22/2022-23:42:53] [E] [TRT] (Unnamed Layer* 0) [Convolution]: count of 6912 weights in kernel, but kernel dimensions (6,6) with 3 input channels, 32 output channels and 1 groups were specified. Expected Weights count is 3 * 66 * 32 / 1 = 3456 [08/22/2022-23:42:53] [E] [TRT] Parameter check failed at: ../builder/Network.cpp::addResize::940, condition: input.getDimensions().nbDims > 0 yolov5ORG: /tensorrtx/yolov5/yolov5.cpp:63: nvinfer1::ICudaEngine build_engine(unsigned int, nvinfer1::IBuilder*, nvinfer1::IBuilderConfig*, nvinfer1::DataType, float&, float&, std::string&): Assertion `upsample11' failed. Aborted (core dumped)

---

Also I try to download yolo5m and yolo5l wts from model zoo there is also errors.

do I need to change code for model m and moe l ?

[wang-xinyu]

Please check the readme. The last cli argument means the model type is s, m or l.

[MyraBaba]

Did you mean : https://github.com/ultralytics/yolov5/releases/download/v6.2/yolov5s-cls.pt

On 22 Aug 2022, at 06:36, Wang Xinyu @.***> wrote:

Did you try the official trained .pt model? https://github.com/wang-xinyu/tensorrtx/tree/master/yolov5#different-versions-of-yolov5 https://github.com/wang-xinyu/tensorrtx/tree/master/yolov5#different-versions-of-yolov5 — Reply to this email directly, view it on GitHub https://github.com/wang-xinyu/tensorrtx/issues/1078#issuecomment-1221759573, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEFRZHZ6ZOOZBIIRJ6XUJH3V2LYTPANCNFSM57DHEOPQ. You are receiving this because you authored the thread.

[wang-xinyu]

Like this:

./yolov5 -s yolov5s.wts yolov5s.engine s
./yolov5 -s yolov5m.wts yolov5m.engine m
./yolov5 -s yolov5l.wts yolov5l.engine l

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.