Converter for aten::floor

[lablabla]

I'm trying to run TRTorch on yolov3 and in the process I added an activation for leaky_relu
Now running trtorch.compile I get an error:

RuntimeError: [enforce fail at core/conversion/conversion.cpp:112] Expected converter to be true but got false
Unable to convert node: %458 : Float() = aten::floor(%2), scope: __module.module_list.85/__module.module_list.85.upsample_85 # /home/devel/anaconda3/envs/tensorrt/lib/python3.6/site-packages/torch/nn/functional.py:2864:0 (conversion.AddLayer)
Schema: aten::floor.int(int a) -> (int)
Converter for aten::floor requested, but no such converter was found.
If you need a converter for this operator, you can try implementing one yourself
or request a converter: https://www.github.com/NVIDIA/TRTorch/issues

but looking at the unary.cpp file in the converters I see that floor is registered:

convert(floor, kFLOOR);

I don't mind implementing it myself but I don't understand why this is happening. I'm guessing that it has something to do with the schema being aten::floor.int(int a) -> (int) and not float?

how would one go about adding a function matching this schema?

I'm using master branch, TensorRT 7.0.1.1

[nmonhollen]

Hi lablabla, I've encountered a similar error when running trtorch.compile for a Fast-SCNN model.

RuntimeError: [enforce fail at core/conversion/conversion.cpp:112] Expected converter to be true but got false Unable to convert node: %376 : Tensor = aten::floor(%324) # /home/nmonhollen/anaconda3/envs/pytorchenv/lib/python3.7/site-packages/torch/nn/functional.py:3010:0 (conversion.AddLayer) Schema: aten::floor.int(int a) -> (int) Converter for aten::floor requested, but no such converter was found. If you need a converter for this operator, you can try implementing one yourself or request a converter: https://www.github.com/NVIDIA/TRTorch/issues

I was wondering what you had found regarding this issue that might be similar to what I'm experiencing?

For reference I'm using:
TensorRT 7.0.0.11
cuDNN 7.6.5
Pytorch 1.6.0
Python 3.6.9

Thanks!