Closed Adarine99 closed 3 years ago
[Convolution]: kernel weights has count 2048 but 4096 was expected it means your layer setup is not right.
one debug method is set output on shadow layers, and comment out following layers. for example, you can set the 3rd layer as output, and check if running fine. and then check 6th layer, then 9th layer.
[Convolution]: kernel weights has count 2048 but 4096 was expected it means your layer setup is not right.
one debug method is set output on shadow layers, and comment out following layers. for example, you can set the 3rd layer as output, and check if running fine. and then check 6th layer, then 9th layer.
got it! Thanks
scaled-yolov4 link:
createEngine as follows: `ICudaEngine createEngine(unsigned int maxBatchSize, IBuilder builder, IBuilderConfig config, DataType dt) { INetworkDefinition network = builder->createNetworkV2(0U);
`
build success;
run error: set name out [12/02/2020-14:17:01] [E] [TRT] (Unnamed Layer 6) [Convolution]: kernel weights has count 2048 but 4096 was expected [12/02/2020-14:17:01] [E] [TRT] (Unnamed Layer 6) [Convolution]: count of 2048 weights in kernel, but kernel dimensions (1,1) with 64 input channels, 64 output channels and 1 groups were specified. Expected Weights count is 64 11 64 / 1 = 4096 [12/02/2020-14:17:01] [E] [TRT] Could not compute dimensions for (Unnamed Layer 6) [Convolution]_output, because the network is not valid [12/02/2020-14:17:01] [E] [TRT] Network validation failed. build out scaled_yolov4:/tensorrtx/yolov4_scaled/scaled_yolov4.cpp:532: void APIToModel(unsigned int, nvinfer1::IHostMemory**): Assertion `engine != nullptr' failed.
Thanks~