|View source on GitHub|
Get a partitioner for VariableScope to keep shards below
Compat aliases for migration
See Migration guide for more details.
tf.variable_axis_size_partitioner( max_shard_bytes, axis=0, bytes_per_string_element=16, max_shards=None )
This partitioner will shard a Variable along one axis, attempting to keep
the maximum shard size below
max_shard_bytes. In practice, this is not
always possible when sharding along only one axis. When this happens,
this axis is sharded as much as possible (i.e., every dimension becomes
a separate shard).
If the partitioner hits the
max_shards limit, then each shard may end up
max_shard_bytes. By default
None and no
limit on the number of shards is enforced.
One reasonable value for
(64 << 20) - 1, or almost
64MB, to keep below the protobuf byte limit.
||The maximum size any given shard is allowed to be.|
||The axis to partition along. Default: outermost axis.|
The maximum number of shards in int created taking precedence
A partition function usable as the
||If any of the byte counts are non-positive.|