Normalizes layer outputs for every minibatch for each output (feature) independently and applies affine transformation to preserve representation of the layer.
op_batch_normalization(operand, scale, bias, running_mean, running_inv_std, spatial, normalization_time_constant = 5000, blent_time_constant = 0, epsilon = 1e-05, use_cudnn_engine = FALSE, name = "", running_count = NULL)
bias | (bool) – whether to include bias |
---|---|
epsilon | (float, default 0.00001) - added to avoid division by 0 |
name | (str) the name of the Function instance in the network |
running_count |