Differences with torch.nn.SyncBatchNorm

torch.nn.SyncBatchNorm

class torch.nn.SyncBatchNorm(
    num_features,
    eps=1e-05,
    momentum=0.1,
    affine=True,
    track_running_stats=True,
    process_group=None
)(input) -> Tensor

For more information, see torch.nn.SyncBatchNorm.

mindspore.nn.SyncBatchNorm

class mindspore.nn.SyncBatchNorm(
    num_features,
    eps=1e-5,
    momentum=0.9,
    affine=True,
    gamma_init='ones',
    beta_init='zeros',
    moving_mean_init='zeros',
    moving_var_init='ones',
    use_batch_statistics=None,
    process_groups=None
)(x) -> Tensor

For more information, see mindspore.nn.SyncBatchNorm.

Differences

PyTorch: Perform cross-device synchronous batch normalization of input data.

MindSpore: MindSpore API is basically the same as PyTorch, and the MindSpore input only supports 2D and 4D. The default value of momentum in MindSpore is 0.9, and the conversion relationship with PyTorch momentum is 1-momentum, with the same default behavior as PyTorch. The training and the parameter update strategy during inference are different from PyTorch.

Categories	Subcategories	PyTorch	MindSpore	Differences
Parameters	Parameter 1	num_features	num_features	-
	Parameter 2	eps	eps	-
	Parameter 3	momentum	momentum	Consistent functionality, but the default value is 0.1 in PyTorch and 0.9 in MindSpore. The conversion relationship with PyTorch momentum is 1-momentum with the same default behavior as PyTorch
	Parameter 4	affine	affine	-
	Parameter 5	track_running_stats	use_batch_statistics	Consistent function. Different values correspond to different default methods.
	Parameter 6	-	gamma_init	PyTorch does not have this parameter, and MindSpore can initialize the value of the parameter gamma
	Parameter 7	-	beta_init	PyTorch does not have this parameter, and MindSpore can initialize the value of the parameter beta
	Parameter 8	-	moving_mean_init	PyTorch does not have this parameter, and MindSpore can initialize the value of the parameter moving_mean
	Parameter 9	-	moving_var_init	PyTorch does not have this parameter, and MindSpore can initialize the value of the parameter moving_var
	Parameter 10	process_group	process_groups	-
Input	Single input	input	x	Interface input. MindSpore only supports 2-D and 4-D input

The detailed differences are as follows: BatchNorm is a special regularization method in the CV field. It has different computation processes during training and inference and is usually controlled by operator attributes. BatchNorm of MindSpore and PyTorch uses two different parameter groups at this point.

• Difference 1

torch.nn.SyncBatchNorm status under different parameters

training	track_running_stats	Status
True	True	Expected training status. running_mean and running_var trace the statistical features of the batch in the entire training process. Each group of input data is normalized based on the mean and var statistical features of the current batch, and then running_mean and running_var are updated.
True	False	Each group of input data is normalized based on the statistics feature of the current batch, but the running_mean and running_var parameters do not exist.
False	True	Expected inference status. The BN uses running_mean and running_var for normalization and does not update them.
False	False	The effect is the same as that of the second status. The only difference is that this is the inference status and does not learn the weight and bias parameters. Generally, this status is not used.

mindspore.nn.SyncBatchNorm status under different parameters

use_batch_statistics	Status
True	Expected training status. moving_mean and moving_var trace the statistical features of the batch in the entire training process. Each group of input data is normalized based on the mean and var statistical features of the current batch, and then moving_mean and moving_var are updated.
Fasle	Expected inference status. The BN uses moving_mean and moving_var for normalization and does not update them.
None	use_batch_statistics is automatically set. For training, set use_batch_statistics to True. For inference, set use_batch_statistics to False.

Compared with torch.nn.SyncBatchNorm, mindspore.nn.SyncBatchNorm does not have two redundant states and retains only the most commonly used training and inference states.

• Difference 2

In PyTorch, the network is in training mode by default, while in MindSpore, it is in inference mode by default (is_training is False). You need to use the net.set_train() method in MindSpore to switch the network to training mode. In this case, the parameters mean and variance are calculated during the training. Otherwise, in inference mode, the parameters are loaded from the checkpoint.

• Difference 3

The meaning of the momentum parameter of the BatchNorm series operators in MindSpore is opposite to that in PyTorch. The relationship is as follows:

\[momentum_{pytorch} = 1 - momentum_{mindspore}\]