Conversion of ONNX to TensorRT on Jetson Xavier

[source821]

Hello,

Please I need help.

I have been struggling trying to convert ONNX model to Tensorrt.

The model was trained in tensorflow version 2.

Below is the log file of the error I am getting with trtexec.

I have also attached the onnx model.

I will deeply appreciate your support.

log (2).txt (509.2 KB)
face-model3-nms.onnx (17.8 MB)

[AastaLLL]

Hi,

[07/29/2021-11:39:44] [V] [TRT] ModelImporter.cpp:125: Slice__567 [Slice] inputs: [NonMaxSuppression__563:0 -> ()], [const_ends__703 -> (1)], [const_ends__504 -> (1)], [const_ends__761 -> (1)],
[07/29/2021-11:39:44] [E] [TRT] onnx_graphsurgeon_node_0: PluginV2Layer must be V2DynamicExt when there are runtime input dimensions.
ERROR: builtin_op_importers.cpp:3040 In function importSlice:
[4] Assertion failed: -r <= axis && axis < r

It looks like the model doesn’t follow the rule for BatchedNMS_TRT.

May I know which TensorFlow model do you use?
And how do you convert the model into ONNX format?

Thanks.