tf.split( value, num_or_size_splits, axis=0, num=None, name='split' )
See the guide: Tensor Transformations > Slicing and Joining
Splits a tensor into sub tensors.
num_or_size_splits is an integer type,
num_split, then splits
num_split smaller tensors.
num_split evenly divides
num_or_size_splits is not an integer type, it is presumed to be a Tensor
size_splits, then splits
len(size_splits) pieces. The shape
i-th piece has the same size as the
value except along dimension
axis where the size is
.html# 'value' is a tensor with shape [5, 30] .html# Split 'value' into 3 tensors with sizes [4, 15, 11] along dimension 1 split0, split1, split2 = tf.split(value, [4, 15, 11], 1) tf.shape(split0) .html# [5, 4] tf.shape(split1) .html# [5, 15] tf.shape(split2) .html# [5, 11] .html# Split 'value' into 3 tensors along dimension 1 split0, split1, split2 = tf.split(value, num_or_size_splits=3, axis=1) tf.shape(split0) .html# [5, 10]
num_or_size_splits: Either a 0-D integer
Tensorindicating the number of splits along split_dim or a 1-D integer
Tensorcontaining the sizes of each output tensor along split_dim. If a scalar then it must evenly divide
value.shape[axis]; otherwise the sum of sizes along the split dimension must match that of the
axis: A 0-D
Tensor. The dimension along which to split. Must be in the range
[-rank(value), rank(value)). Defaults to 0.
num: Optional, used to specify the number of outputs when it cannot be inferred from the shape of
name: A name for the operation (optional).
num_or_size_splits is a scalar returns
num_or_size_splits is a 1-D Tensor returns
Tensor objects resulting from splitting
numis unspecified and cannot be inferred.