Errors when Converter to TensorRT by torch2trt

Hi all:

I try to converter model to TensorRT by torch2trt,

and return error message:

[01/31/2024-10:51:39] [TRT] [W] Tensor DataType is determined at build time for tensors not marked as input or output.
[01/31/2024-10:51:46] [TRT] [E] 10: Could not find any implementation for node decoder.blocks.3.conv1.0:0:CONVOLUTION:GPU + decoder.blocks.3.conv1.1:0:SCALE:GPU + decoder.blocks.3.conv1.2:0:RELU:GPU.
[01/31/2024-10:51:46] [TRT] [E] 10: [optimizer.cpp::computeCosts::3869] Error Code 10: Internal Error (Could not find any implementation for node decoder.blocks.3.conv1.0:0:CONVOLUTION:GPU + decoder.blocks.3.conv1.1:0:SCALE:GPU + decoder.blocks.3.conv1.2:0:RELU:GPU.)
encoder=resnet34,W=1056,H=1056,Batch Size=1,rtMode=8
Traceback (most recent call last):
File “/home/OrinNX/AI_Python/TranModelToTensorRT.py”, line 200, in
ai_model_obj.Convert_RT(rt_model_path=rt_model_path, max_trt_batch_size=predict_batch_size,device=device )
File “/home/OrinNX/AI_Python/SegSMP.py”, line 1776, in Convert_RT
output_trt = self.model_trt(im)
File “/home/OrinNX/.local/lib/python3.10/site-packages/torch/nn/modules/module.py”, line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File “/home/OrinNX/.local/lib/python3.10/site-packages/torch/nn/modules/module.py”, line 1527, in _call_impl
return forward_call(*args, **kwargs)
File “/home/OrinNX/AI_Python/torch2trt/torch2trt.py”, line 614, in forward
idx = self.engine.get_binding_index(input_name)
AttributeError: ‘NoneType’ object has no attribute ‘get_binding_index’

How can I fix it?

Thanks.

Hi,

This is a known issue in JetPack 6 DP.
Please wait for our GA release for the fix.

Thanks.

1 Like

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.