A primitive to perform batch normalization. More...

Functions
dnnl_status_t DNNL_API	dnnl_batch_normalization_forward_desc_init (dnnl_batch_normalization_desc_t bnrm_desc, dnnl_prop_kind_t prop_kind, const dnnl_memory_desc_t data_desc, float epsilon, unsigned flags)
	Initializes a batch normalization descriptor `bnrm_desc` for forward propagation using `prop_kind` (possible values are dnnl_forward_training and dnnl_forward_inference), memory descriptor `data_desc`, normalization parameter `epsilon`, and `flags` set using bit flags of type dnnl_batch_normalization_desc_t. More...

dnnl_status_t DNNL_API	dnnl_batch_normalization_backward_desc_init (dnnl_batch_normalization_desc_t bnrm_desc, dnnl_prop_kind_t prop_kind, const dnnl_memory_desc_t diff_data_desc, const dnnl_memory_desc_t *data_desc, float epsilon, unsigned flags)
	Initializes a batch normalization descriptor `bnrm_desc` for backward propagation with respect to data and scale-shift parameters using memory descriptors `data_desc` and `diff_data_desc`, normalization parameter `epsilon`, and `flags` set using bit flags of type dnnl_batch_normalization_desc_t. More...

Detailed Description

A primitive to perform batch normalization.

Both forward and backward passes support in-place operation; that is, src and dst point to the same memory for forward pass, and diff_dst and diff_src point to the same memory for backward pass.

Batch normalization supports different flavors controlled by dnnl_batch_normalization_desc_t. For example, batch normalization can compute the mean and variance on its own or take them as inputs. It can either perform scaling and shifting using gamma and beta parameters or not. Optionally, it can also perform a fused ReLU, which in case of training would also require a workspace.

See also: dnnl_batch_normalization_desc_t; Batch Normalization in developer guide; Batch normalization in C++ API

Function Documentation

◆ dnnl_batch_normalization_forward_desc_init()

dnnl_status_t DNNL_API dnnl_batch_normalization_forward_desc_init	(	dnnl_batch_normalization_desc_t *	bnrm_desc,
		dnnl_prop_kind_t	prop_kind,
		const dnnl_memory_desc_t *	data_desc,
		float	epsilon,
		unsigned	flags
	)

Initializes a batch normalization descriptor bnrm_desc for forward propagation using prop_kind (possible values are dnnl_forward_training and dnnl_forward_inference), memory descriptor data_desc, normalization parameter epsilon, and flags set using bit flags of type dnnl_batch_normalization_desc_t.

Inputs:

src (dnnl_query_src_md, 0)
mean (dnnl_query_src_md, 1), if dnnl_use_global_stats bit-flags is set in flags
variance (dnnl_query_src_md, 2), if dnnl_use_global_stats bit-flags is set in flags
scale_and_shift (dnnl_query_weights_md, 0), if dnnl_use_scaleshift bit-flags is set in flags

Outputs:

dst (dnnl_query_dst_md, 0)
mean (dnnl_query_dst_md, 1), if dnnl_use_global_stats bit-flags is not set in flags prop_kind = dnnl_forward_training
variance (dnnl_query_dst_md, 2), if dnnl_use_global_stats bit-flags is not set in flags and prop_kind = dnnl_forward_training
workspace (dnnl_query_workspace_md, 0), if dnnl_fuse_norm_relu bit-flags is set in flags and prop_kind = dnnl_forward_training

Note: In-place operation is supported; that is, dst points to the same memory as src.

See also: dnnl_batch_normalization_desc_t

◆ dnnl_batch_normalization_backward_desc_init()

dnnl_status_t DNNL_API dnnl_batch_normalization_backward_desc_init	(	dnnl_batch_normalization_desc_t *	bnrm_desc,
		dnnl_prop_kind_t	prop_kind,
		const dnnl_memory_desc_t *	diff_data_desc,
		const dnnl_memory_desc_t *	data_desc,
		float	epsilon,
		unsigned	flags
	)

Initializes a batch normalization descriptor bnrm_desc for backward propagation with respect to data and scale-shift parameters using memory descriptors data_desc and diff_data_desc, normalization parameter epsilon, and flags set using bit flags of type dnnl_batch_normalization_desc_t.