RuntimeError: Expected tensor for argument #1 'indices' to have scalar type Long; but got CPUFloatTensor instead (while checking arguments for embedding)

Right now I’m trying to replicated the input dimension size of a loaded model to convert the content into ONNX. Initially I was given an error for not having enough input arguments for any people who remember my previous post (Missing key(s) in state_dict - #2 by ptrblck). I fixed my previous error and now facing another one.

I think it has something to do with my forward function handling the data type.

partThree

what is the datatype of the input tensor for embedding ?