brainpy.math.sharding: Parallelization Support

device_mesh
get_sharding	Get sharding according to the given axes information.
partition_by_axname	Put the given arrays into the mesh devices.
partition_by_sharding	Partition inputs with the given sharding strategy.
partition	Partition the input arrays onto devices by the given sharding strategies.
keep_constraint	Keep the sharding constraint of the given inputs during computation.

The commonly used axis names.

NEU_AXIS	str(object='') -> str str(bytes_or_buffer[, encoding[, errors]]) -> str
PRE_AXIS	str(object='') -> str str(bytes_or_buffer[, encoding[, errors]]) -> str
POST_AXIS	str(object='') -> str str(bytes_or_buffer[, encoding[, errors]]) -> str
SYN_AXIS	str(object='') -> str str(bytes_or_buffer[, encoding[, errors]]) -> str
TIME_AXIS	str(object='') -> str str(bytes_or_buffer[, encoding[, errors]]) -> str
BATCH_AXIS	str(object='') -> str str(bytes_or_buffer[, encoding[, errors]]) -> str