Cast (CPU)
==========

* Domain: ``ai.onnx``
* Device: ``CPU``
* Kernel: ``onnx_light_cpu::Cast``
* Supported types: ``FLOAT``, ``DOUBLE``, ``FLOAT16``, ``BFLOAT16``, ``INT8``, ``UINT8``, ``INT16``, ``UINT16``, ``INT32``, ``UINT32``, ``INT64``, ``UINT64``, ``BOOL``, ``STRING``, ``FLOAT8E4M3FN``, ``FLOAT8E4M3FNUZ``, ``FLOAT8E5M2``, ``FLOAT8E5M2FNUZ``, ``FLOAT8E8M0``, ``INT4``, ``UINT4``, ``INT2``, ``UINT2``, ``FLOAT4E2M1``, ``FLOAT6E2M3``, ``FLOAT6E3M2``, ``COMPLEX64``, ``COMPLEX128``
* Opset bounds: none

Operator support
----------------

The standard ONNX schema and any standard shape inference come from onnx-light. No additional shape, peak-memory, fusion-pattern, or gradient implementation is registered by onnx-light-cpu for this operator.