tf.scatter_max(
ref,
indices,
updates,
use_locking=False,
name=None
)
Defined in generated file: tensorflow/python/ops/gen_state_ops.py
.
Reduces sparse updates into a variable reference using the max
operation.
This operation computes
# Scalar indices
ref[indices, ...] = max(ref[indices, ...], updates[...])
# Vector indices (for each i)
ref[indices[i], ...] = max(ref[indices[i], ...], updates[i, ...])
# High rank indices (for each i, ..., j)
ref[indices[i, ..., j], ...] = max(ref[indices[i, ..., j], ...], updates[i, ..., j, ...])
This operation outputs ref
after the update is done.
This makes it easier to chain operations that need to use the reset value.
Duplicate entries are handled correctly: if multiple indices
reference
the same location, their contributions combine.
Requires updates.shape = indices.shape + ref.shape[1:]
or updates.shape = []
.
Args:
ref
: A mutableTensor
. Must be one of the following types:half
,bfloat16
,float32
,float64
,int32
,int64
. Should be from aVariable
node.indices
: ATensor
. Must be one of the following types:int32
,int64
. A tensor of indices into the first dimension ofref
.updates
: ATensor
. Must have the same type asref
. A tensor of updated values to reduce intoref
.use_locking
: An optionalbool
. Defaults toFalse
. If True, the update will be protected by a lock; otherwise the behavior is undefined, but may exhibit less contention.name
: A name for the operation (optional).
Returns:
A mutable Tensor
. Has the same type as ref
.