* Copyright (c) 2025 Huawei Technologies Co., Ltd.
* This program is free software, you can redistribute it and/or modify it under the terms and conditions of
* CANN Open Software License Agreement Version 2.0 (the "License")
* Please refer to the License for details. You may not use this file except in compliance with the License.
* THIS SOFTWARE IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND, EITHER EXPRESS OR IMPLIED,
* INCLUDING BUT NOT LIMITED TO NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE.
* See LICENSE in the root of the software repository for the full text of the License.
*/
* \file cast_proto.h
* \brief
*/
#ifndef OP_PROTO_CAST_PROTO_H_
#define OP_PROTO_CAST_PROTO_H_
#include "graph/operator_reg.h"
namespace ge
{
*@brief Cast a tensor form src data type to dst data type.
*@par Inputs:
*One input:
* x:A ND or 5HD tensor. Support 1D~8D. Must be one of the following types: bool, float16, float, int8, int32, uint32, uint8, bfloat16, uint1,
int64, uint64, int16, uint16, double, complex32, complex64, complex128, qint8, quint8, qint16, quint16, qint32,
hifloat8, float8_e5m2, float8_e4m3fn, float4_e1m2, float4_e2m1.
*@par Attributes:
*dst_type: A required attribute of type int32, specifying the dst data type.
*@par Outputs:
*y:A ND Tensor with same shape as x, and data type is specified by dst_type.
*@attention Constraints:
* @li In the scenario where the data type is converted from float16 to int16: \n
* If the input data contains inf, inf is converted into the maximum value of int16. \n
* If the input data contains -inf, -inf is converted into the minimum value of int16. \n
* @li In the scenarios where the data type is converted from INT32 to INT8: \n
* It can only guarantee that the input data has no precision errors within the range of (-2048, 1920).
* @li Atlas Inference Series Product in the scenarios where the data type is converted from FLOAT32 to INT8: \n
* It can only guarantee that the input data has no precision errors within the range of (-2048, 1920).
* @li Atlas Inference Series Product in the scenarios where the data type is converted from FLOAT32 to INT64 and from FLOAT32 to UINT8: \n
* It can only guarantee that the input data has no precision errors within the range of (-2147483648, 2147483583).
* @li Atlas Inference Series Product in the scenarios where the data type is converted from INT64 to FLOAT32: \n
* It can only guarantee that the input data has no precision errors within the range of (-2147483648, 2147483647).
* @li Ascend 950 AI Processor in the scenario where the data type is converted from INT32 to INT4: \n
* The last dim of x must be an even number.
*/
REG_OP(Cast)
.INPUT(x, TensorType({DT_BOOL, DT_FLOAT16, DT_FLOAT, DT_INT8, DT_INT32, DT_UINT32, DT_UINT8,
DT_INT64, DT_UINT64, DT_INT16, DT_UINT16, DT_DOUBLE, DT_COMPLEX64,
DT_COMPLEX128, DT_QINT8, DT_QUINT8, DT_QINT16, DT_QUINT16, DT_QINT32, DT_BF16, DT_UINT1,
DT_COMPLEX32, DT_HIFLOAT8, DT_FLOAT8_E5M2, DT_FLOAT8_E4M3FN,
DT_FLOAT4_E1M2, DT_FLOAT4_E2M1}))
.OUTPUT(y, TensorType({DT_BOOL, DT_FLOAT16, DT_FLOAT, DT_INT8, DT_INT32, DT_UINT32, DT_UINT8,
DT_INT64, DT_UINT64, DT_INT16, DT_UINT16, DT_DOUBLE, DT_COMPLEX64,
DT_COMPLEX128, DT_QINT8, DT_QUINT8, DT_QINT16, DT_QUINT16, DT_QINT32,
DT_BF16, DT_COMPLEX32, DT_HIFLOAT8, DT_FLOAT8_E5M2, DT_FLOAT8_E4M3FN,
DT_FLOAT4_E1M2, DT_FLOAT4_E2M1, DT_INT4}))
.REQUIRED_ATTR(dst_type, Int)
.OP_END_FACTORY_REG(Cast)
}
#endif
