[relay]What does the mistake mean? “in particular dimensio 2 conflicts 3 does not match 2

Questions
#1

I try to transform the NLP model using from_tensorflow, then get an error.

......
%1817 = add(%35, %1816);
%1818 = reshape(%1817, newshape(-1,[28500]))
%1819 = topk(%1818, k=3, dtype='int32')
%1820 = %1819.1
%1821 = cast(%1820, dtype='float64')
%1822 = cast(9500, dtype='float64')
%1823 = divide(%1821, %1822)
%1824 = cast(%1823, dtype='int32')
%1825 = add(%33, %1824)
%1826 = reshape(%1825, newshape=[-1])
%1827 = take(%30, %1826, axis=0) in particular dimension 2 conflicts (int64)3 does not match 2; unable to unify: `Tensor[(15, 8, 3, 64), float32]` and `Tensor[(15, 8, 2, 64), float32]`; ;
%1818 = take(%loop_var172, %loop_var50, axis =0)
........

I print same information, looks right.

%30 => shape: (15, 8, 3, 64) op: ConcatV2
%1826 => shape: (15,) op: Reshape
%1827 => node:
name: "while/GatherV2"
input: "while/concat"
input: "while/Reshape_75"
inpurt: "while/GatherV2/axis"
attr{
    key: "Taxis"
    value{
    type: DT_INT32
    }
}
attr{
    key: "Tparams"
    value{
    type: DT_FLOAT
    }
}
attr{
    key: "_output_shapes"
    value{
        list{
           dim {
               size: -1
            }
           dim {
               size: 8
            }
           dim {
               size: -1
            }
           dim {
               size: 64
            }
        }
    }
}
attr{
    key: "batch_dims"
    value{
     i: 0
    }
}

I don’t know what this mistake means and how do I start checking it ?Any good Suggestions and methods to check errors?

Support for tensorflow ops for newer TF version
#2

I find the ‘while/GatherV2’ is the input of ‘NextIteration’,It will continue to be passed in the 'while_loop‘.

name: "while/NextIteration_26"
op: "NextIteration"
input: "while/GatherV2"
.......

name: "while/Merge_26"
op: "Merge"
input: "while/Enter_26"
input: "while/NextIteration_26"
......

But the shape looks right. They are all (-1, 8, -1, 64) and I don’t see (15, 8, 2, 64)…

#3

this must be problem with shape of array where it can’t reshape to target shape, may be you can attached model i can have a look into that

#4

My model is in the Intranet and cannot be taken out.Any good Suggestions and methods to check errors?

#5

better idea is to deep and check where the error is occuring, for this you can use PYCHARM and debug inside the TVM code