Closed gabrer closed 5 years ago
The initialization function of the SeparatedBatchNorm1d module has two arguments eps and momentum.
eps
momentum
def __init__(self, num_features, max_length, eps=1e-5, momentum=0.1, affine=True):
Is this momentum the same we use in the in our optimization algorithm (e.g. SGD), or it's an additional momentum just for the batch normalization process?
I couldn't find any mention about this in the original paper.
@gabrer You can find info in the document of batch_norm http://pytorch.org/docs/master/nn.html#torch.nn.BatchNorm1d
Thank you! I have already found it! :)
The initialization function of the SeparatedBatchNorm1d module has two arguments
eps
andmomentum
.Is this
momentum
the same we use in the in our optimization algorithm (e.g. SGD), or it's an additional momentum just for the batch normalization process?I couldn't find any mention about this in the original paper.