* 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 quantize.h
* \brief Quantize is an Quantization function which refers to
* techniques for performing computations and storing tensors at lower bitwidths than floating point precision.
* Mathematical formulas: Quantize(x, scale, offset) = (x* scale) + offset
*/
#ifndef LIB_QUANTIZATION_QUANTIZE_H
#define LIB_QUANTIZATION_QUANTIZE_H
#if defined(__NPU_ARCH__) && (__NPU_ARCH__ == 3101 || __NPU_ARCH__ == 5102)
#include "kernel_tensor.h"
#include "include/adv_api/quantization/quantize_utils.h"
#include "../../../impl/adv_api/detail/common/check.h"
#include "../../../impl/adv_api/detail/quantization/quantize/quantize_impl.h"
namespace AscendC {
#pragma begin_pipe(V)
* \ingroup Quantize
* \brief Quantize interface
* \tparam config: template parameter, see detail at QuantizeConfig
* \tparam DstT: dstTensor data type
* \tparam SrcT: srcTensor data type
* \tparam ScaleT: offset data type, can be a scalar or LocalTensor
* \tparam OffsetT: offset data type, can be a scalar or LocalTensor
* \param [out] dstTensor: output LocalTensor
* \param [in] srcTensor: input LocalTensor
* \param [in] scale: quantization parameter, tensor or scalar
* \param [in] offset: quantization parameter, tensor or scalar
* \param [in] params: quantization parameter, see detail at QuantizeParams
*/
template <const QuantizeConfig& config, typename DstT, typename SrcT, typename ScaleT, typename OffsetT>
__aicore__ inline void Quantize(const LocalTensor<DstT>& dstTensor, const LocalTensor<SrcT>& srcTensor,
const ScaleT& scale, const OffsetT& offset, const QuantizeParams& params)
{
QuantizeImpl<config, DstT, SrcT, ScaleT, OffsetT>(dstTensor, srcTensor, scale, offset, params);
}
* \ingroup Quantize
* \brief Quantize interface
* \tparam config: template parameter, see detail at QuantizeConfig
* \tparam DstT: dstTensor data type
* \tparam SrcT: srcTensor data type
* \tparam ScaleT: offset data type, can be a scalar or LocalTensor
* \tparam OffsetT: offset data type, can be a scalar or LocalTensor
* \param [out] dstTensor: output LocalTensor
* \param [in] srcTensor: input LocalTensor
* \param [in] sharedTmpBuffer：extra temporary shared space used for intermediate values among calculation process,
* whose required space size should refer to corresponding tiling API, which is defined at
* quantize_tiling.h. Generally, the more space you allocate, the better performance you will achieve,
* and the performance reaches peak when buffer size is maximum(calculated by tiling function). Moreover, it
* is not guaranteed that the shared space will be cleared after usage, the data could be anything.
* \param [in] scale: quantization parameter, tensor or scalar
* \param [in] offset: quantization parameter, tensor or scalar
* \param [in] params: quantization parameter, see detail at QuantizeParams
*/
template <const QuantizeConfig& config, typename DstT, typename SrcT, typename ScaleT, typename OffsetT>
__aicore__ inline void Quantize(const LocalTensor<DstT>& dstTensor, const LocalTensor<SrcT>& srcTensor,
const LocalTensor<uint8_t>& sharedTmpBuffer, const ScaleT& scale, const OffsetT& offset, const QuantizeParams& params)
{
CheckTensorPosition(sharedTmpBuffer, "sharedTmpBuffer", "VECIN, VECOUT, VECCALC");
QuantizeImpl<config, DstT, SrcT, ScaleT, OffsetT>(dstTensor, srcTensor, scale, offset, params);
}
#pragma end_pipe
}
#endif
#endif
