AttributeError: 'tensorrt_bindings.tensorrt.INetworkDefinition' object has no attribute 'add_convolution'. Did you mean: 'add_convolution_nd'?

[AWallyAllah]

I'm converting a PyTorch model with MobileNetV2 backbone, However I get this error:

AttributeError: 'tensorrt_bindings.tensorrt.INetworkDefinition' object has no attribute 'add_convolution'. Did you mean: 'add_convolution_nd'?

Versions:

• tensorrt==10.0.0b6
• torch2trt==0.4.0
• CUDA 11.5, V11.5.119
• Python 3.10.12

[xmdszzz]

@AWallyAllah hey bro, i checked the latest release notes, v0.4.0 only supports up to TensorRT 8, so TensorRT10 is not supported, and you should change the TRT's version.

[jaybdub]

Hi @AWallyAllah , @xmdszzz,

We've just added support for TensorRT 10 in the master branch.

Could you try this?

Let me know if you run into any issues or this works for you.

Best,
John