Dimensionality of oracle

[yiyanggong]

Hi, I'm running the code, and it's throwing an error when compiling the model for the oracles. The neg_loss function is expecting a two element y_pred for each gfp variant, but the model for the oracles generated by the function build_pred_model only outputs a one element y_pred. The full error output when running [run_gfp_tests.py] is below:

ValueError: slice index 1 of dimension 1 out of bounds. for 'loss/dense_1_loss/strided_slice_2' (op: 'StridedSlice') with input shapes: [?,1], [2], [2], [2] and with computed input tensors: input[1] = <0 1>, input[2] = <0 2>, input[3] = <1 1>.

Thanks.

[SimonBartels]

I have the same problem.

[SimonBartels]

It seems that in the original code the dense layer in build_pred_model was of size 2 and not 1.