Created by: bliunlpr

When I print the paramters of the model, I found: _batch_norm_2.w0 (1, 3072) _batch_norm_2.w1 (1, 3072) _batch_norm_2.w2 (1, 3072) _batch_norm_2.wbias (1, 3072). I wonder why batch-norm layer has four paramters(w0, w1, w2, wbias)? I guess these are \gamma, \beta, running_average, running_var. So what's the meaning for batch-norm paramters respectively? Thanks very much!