size of mapping function? about wb_srgb

[mahmoudnafifi]

For each training image and the corresponding ground truth image, we computed a 3x11 matrix (in the code we use the transposed version of it and it is represented as an 11x3 matrix, then we vectorized it to be a 33-d vector). That is why our mapping functions are stored as an nx33 matrix, where n is the total number of training data.
In the paper, we used WB_model.mat (or mappingFuncs.npy for Python).
I think you used WB_model+.mat (or mappingFuncs+.npy for Python) which is an upgraded model with new training examples, as explained here https://github.com/mahmoudnafifi/WB_sRGB/blob/master/README.md