jax.lax.dot_general

jax.lax.dot_general(lhs, rhs, dimension_numbers, precision=None, preferred_element_type=None)

General dot product/contraction operator.

Wraps XLA’s DotGeneral operator.

The semantics of dot_general are complicated, but most users should not have to use it directly. Instead, you can use higher-level functions like jax.numpy.dot(), jax.numpy.matmul(), jax.numpy.tensordot(), jax.numpy.einsum(), and others which will construct appropriate calls to dot_general under the hood. If you really want to understand dot_general itself, we recommend reading XLA’s DotGeneral operator documentation.

Parameters

• lhs (Union[Array, ndarray, bool_, number, bool, int, float, complex]) – an array

• rhs (Union[Array, ndarray, bool_, number, bool, int, float, complex]) – an array

• dimension_numbers (Tuple[Tuple[Sequence[int], Sequence[int]], Tuple[Sequence[int], Sequence[int]]]) – a tuple of tuples of sequences of ints of the form ((lhs_contracting_dims, rhs_contracting_dims), (lhs_batch_dims, rhs_batch_dims))

• precision (Union[None, str, Precision, Tuple[str, str], Tuple[Precision, Precision]]) – Optional. Either None, which means the default precision for the backend, a Precision enum value (Precision.DEFAULT, Precision.HIGH or Precision.HIGHEST) or a tuple of two Precision enums indicating precision of lhs` and rhs.

• preferred_element_type (Union[Any, str, dtype, SupportsDType, None]) – Optional. Either None, which means the default accumulation type for the input types, or a datatype, indicating to accumulate results to and return a result with that datatype.

Return type

Array

Returns

An array whose first dimensions are the (shared) batch dimensions, followed by the lhs non-contracting/non-batch dimensions, and finally the rhs non-contracting/non-batch dimensions.