Interfaces for extensions, selectively lifted out of
class Block: A representation of a block of code
in TFF's internal language.
class Call: A representation of a function
invocation in TFF's internal language.
class CompiledComputation: A
representation of a fully constructed and serialized computation.
The abstract base class for abstractions in the TFF's internal language.
class Intrinsic: A representation of an
intrinsic in TFF's internal language.
class Lambda: A representation of a lambda
expression in TFF's internal language.
class Placement: A representation of a
placement literal in TFF's internal language.
class Reference: A reference to a name
defined earlier in TFF's internal language.
class Selection: A selection by name or index
from a tuple-typed value in TFF's language.
class TFParser: Callable taking subset of TFF
AST constructs to CompiledComputations.
class Tuple: A tuple with named or unnamed
elements in TFF's internal language.
Converts a computation building block to a computation impl.
Checks whitelist of intrinsics reducible to aggregate or broadcast.
Creates a called federated map of equal values.
Creates a called federated map or apply depending on
a called federated zip.
Gets a Python
dict of the unbound references in
the block variables in
comp whitelisted by
Inserts an identity TF graph called on References under
target_type is assignable from
comp is a called intrinsic with the
type_spec against an explicit whitelist for
Merges all the tuples of intrinsics in
comp into one intrinsic.
Removes all the mapped or applied identity functions in
Replaces all the called lambdas in
comp with a block.
Replaces any selection from a tuple with the underlying tuple element.
comp recursively postorder and replaces its constituents.
Walks type tree of
type_signature postorder, calling
Returns nested structure of
tf.TensorSpecs for a given TFF type.
Yields a new unique name that does not exist in
Replaces all the bound reference names in
comp with unique names.
placement, returning a single call to map, apply or value.