count_normalization is only correct for batch_norm. wrong flops count for layernorm

[pzpzpzp2]

pytorch-OpCounter/thop/profile.py

Line 32 in 43c064a

nn.LayerNorm: count_normalization,

The same count_normalization function is used for every norm-esque module but batchnorms store an estimate mean and stdev, while layernorms calculate them at inference time. Shouldn't layernorms account for the cost of evaluating mean and stdev? The difference is pretty significant:
The mean is n flops, stdev is 2n more flops? and thats before the rest of the norm module which is another 2n.
Is there a reason layernorms should be estimateable as only 2n flops by re-using batchnorm's estimate?

[dyhBUPT]

I think you are right.
BTW, I think the MACs of BN (eval, no affine) should be n, not 2n in codes.

pytorch-OpCounter/thop/vision/basic_hooks.py

Lines 60 to 69 in 43c064a

	def count_normalization(m: nn.modules.batchnorm._BatchNorm, x, y):
	# TODO: add test cases
	# https://github.com/Lyken17/pytorch-OpCounter/issues/124
	# y = (x - mean) / sqrt(eps + var) * weight + bias
	x = x[0]
	# bn is by default fused in inference
	flops = calculate_norm(x.numel())
	if (getattr(m, 'affine', False) or getattr(m, 'elementwise_affine', False)):
	flops *= 2
	m.total_ops += flops

pytorch-OpCounter/thop/vision/calc_func.py

Lines 43 to 45 in 43c064a

	def calculate_norm(input_size):
	"""input is a number not a array or tensor"""
	return torch.DoubleTensor([2 * input_size])