Source code for einx._src.frontend.impl.tensorflow

import einx._src.tracer as tracer
import einx._src.adapter as adapter
from ..api import api
from ..types import Tensor
import types
import inspect
import functools
from functools import partial
from ..backend import registry
from ..backend import Backend
from ._util import _make_iskwarg
from ._docs import _make_doc_adapt_numpylike_reduce
from ._docs import _make_doc_adapt_numpylike_elementwise


def _get_backend_kwargs():
    tf = tracer.signature.python.import_("tensorflow", as_="tf")
    optimizations = [
        tracer.optimizer.classical.SkipReshape(tf.reshape),
        tracer.optimizer.classical.SkipTranspose(tf.transpose),
        tracer.optimizer.classical.SkipBroadcastTo(tf.broadcast_to),
        tracer.optimizer.classical.SkipConcatenate(tf.concat),
        tracer.optimizer.InlineGraph(),
        tracer.optimizer.SkipCast(),
    ]

    import tensorflow as tf

    def is_supported_tensor(tensor):
        return isinstance(tensor, tf.Tensor)

    def get_shape(tensor):
        return tuple(int(x) for x in tensor.shape)

    return {
        "optimizations": optimizations,
        "compiler": tracer.compiler.python,
        "is_supported_tensor": is_supported_tensor,
        "get_shape": get_shape,
        "wrap_construct_graph": lambda x: x,
    }




def adapt_numpylike_reduce(op):
    iskwarg = lambda name, iskwarg=_make_iskwarg(op): name != "axis" and iskwarg(name)
    tf = tracer.signature.tensorflow()

    classical = adapter.classical_from_tensorflow.ops(tf)

    op = tracer.signature.python.constant(op)
    op = adapter.decomposednamedtensor_from_classical.reduce(op, expected_type=tf.Tensor)
    op = adapter.namedtensor_from_decomposednamedtensor.op(op, classical)
    op = adapter.namedtensor_calltensorfactory.op(op, expected_type=tf.Tensor)
    op = adapter.einx_from_namedtensor.reduce(op, iskwarg=iskwarg)

    return api(op, backend=types.SimpleNamespace(**_get_backend_kwargs()))



adapt_numpylike_reduce.__doc__ = _make_doc_adapt_numpylike_reduce()




def adapt_numpylike_elementwise(op):
    iskwarg = lambda name, iskwarg=_make_iskwarg(op): name != "axis" and iskwarg(name)
    tf = tracer.signature.tensorflow()

    classical = adapter.classical_from_tensorflow.ops(tf)

    op = tracer.signature.python.constant(op)
    op = adapter.decomposednamedtensor_from_classical.elementwise(op, classical, expected_type=tf.Tensor)
    op = adapter.namedtensor_from_decomposednamedtensor.op(op, classical)
    op = adapter.namedtensor_calltensorfactory.op(op, expected_type=tf.Tensor)
    op = adapter.einx_from_namedtensor.elementwise(op, iskwarg=iskwarg)

    return api(op, backend=types.SimpleNamespace(**_get_backend_kwargs()))



adapt_numpylike_elementwise.__doc__ = _make_doc_adapt_numpylike_elementwise()


def _backend_creator(create):
    def new_create():
        tf = tracer.signature.tensorflow()

        classical = adapter.classical_from_tensorflow.ops(tf)

        decomposednamedtensor_ops, name = create(tf, classical)

        namedtensor_ops = adapter.namedtensor_from_decomposednamedtensor.ops(decomposednamedtensor_ops, classical)
        namedtensor_ops = adapter.namedtensor_calltensorfactory.ops(namedtensor_ops, expected_type=tf.Tensor)
        einx_ops = adapter.einx_from_namedtensor.ops(namedtensor_ops)

        return Backend(ops=einx_ops, name=name, priority=-5, **_get_backend_kwargs())

    return new_create


@_backend_creator
def create_backend_numpylike(tf, classical):
    decomposednamedtensor_ops = (
        {name: adapter.decomposednamedtensor_from_classical.elementwise(getattr(classical, name), classical) for name in adapter.ops.elementwise}
        | {name: adapter.decomposednamedtensor_from_classical.reduce(getattr(classical, name)) for name in adapter.ops.reduce}
        | {name: adapter.decomposednamedtensor_from_classical.preserve_shape(getattr(classical, name)) for name in adapter.ops.preserve_shape}
        | {name: adapter.decomposednamedtensor_from_classical.argfind(getattr(classical, name), classical) for name in adapter.ops.argfind}
        | {name: adapter.decomposednamedtensor_from_classical.update_at_ravelled(getattr(classical, name), classical) for name in adapter.ops.update_at}
        | {
            "get_at": adapter.decomposednamedtensor_from_classical.get_at_ravelled(classical),
            "dot": adapter.decomposednamedtensor_from_classical.dot(classical),
        }
    )

    return decomposednamedtensor_ops, "tensorflow.numpylike"


registry.register_on_import("tensorflow", "tensorflow.numpylike", create_backend_numpylike)


@_backend_creator
def create_backend_einsum(tf, classical):
    einsum = adapter.einsum_from_tensorflow(tf)

    decomposednamedtensor_ops = {
        "id": adapter.decomposednamedtensor_from_einsum.id(einsum),
        "sum": adapter.decomposednamedtensor_from_einsum.sum(einsum),
        "multiply": adapter.decomposednamedtensor_from_einsum.multiply(einsum),
        "dot": adapter.decomposednamedtensor_from_einsum.dot(einsum),
    }

    return decomposednamedtensor_ops, "tensorflow.einsum"


registry.register_on_import("tensorflow", "tensorflow.einsum", create_backend_einsum)


def create_backend():
    einx_ops_numpylike = create_backend_numpylike().ops
    einx_ops_einsum = create_backend_einsum().ops
    einx_ops = einx_ops_numpylike | {"dot": einx_ops_einsum["dot"]}

    return Backend(ops=einx_ops, name="tensorflow", priority=0, **_get_backend_kwargs())


registry.register_on_import("tensorflow", "tensorflow", create_backend)