pub fn einsum_native_tensor_reads(
operands: &[(&NativeTensorReadInput<'_>, &[usize])],
output_ids: &[usize],
) -> Result<Tensor>Expand description
Execute a cached einsum over read-only native tensor inputs.
Backends may consume borrowed host views directly or materialize/upload them inside their execution session. Mixed dtypes are promoted by materializing only the operands that require conversion.