| /* | |
| * SPDX-FileCopyrightText: Copyright 2010-2018, 2022 Arm Limited and/or its affiliates <open-source-office@arm.com> | |
| * | |
| * SPDX-License-Identifier: Apache-2.0 | |
| * | |
| * Licensed under the Apache License, Version 2.0 (the License); you may | |
| * not use this file except in compliance with the License. | |
| * You may obtain a copy of the License at | |
| * | |
| * www.apache.org/licenses/LICENSE-2.0 | |
| * | |
| * Unless required by applicable law or agreed to in writing, software | |
| * distributed under the License is distributed on an AS IS BASIS, WITHOUT | |
| * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | |
| * See the License for the specific language governing permissions and | |
| * limitations under the License. | |
| */ | |
| /* ---------------------------------------------------------------------- | |
| * Project: CMSIS NN Library | |
| * Title: arm_softmax_q15.c | |
| * Description: Q15 softmax function | |
| * | |
| * $Date: 4 Aug 2022 | |
| * $Revision: V.1.0.2 | |
| * | |
| * Target Processor: Cortex-M cores | |
| * | |
| * -------------------------------------------------------------------- */ | |
| /** | |
| * @ingroup groupNN | |
| */ | |
| /** | |
| * @addtogroup Softmax | |
| * @{ | |
| */ | |
| /* | |
| * Q15 softmax function | |
| * | |
| * | |
| * Here, instead of typical e based softmax, we use | |
| * 2-based softmax, i.e.,: | |
| * | |
| * y_i = 2^(x_i) / sum(2^x_j) | |
| * | |
| * The relative output will be different here. | |
| * But mathematically, the gradient will be the same | |
| * with a log(2) scaling factor. | |
| * | |
| */ | |
| void arm_softmax_q15(const q15_t *vec_in, const uint16_t dim_vec, q15_t *p_out) | |
| { | |
| q31_t sum; | |
| int16_t i; | |
| uint8_t shift; | |
| q31_t base; | |
| base = -1 * 0x100000; | |
| for (i = 0; i < dim_vec; i++) | |
| { | |
| if (vec_in[i] > base) | |
| { | |
| base = vec_in[i]; | |
| } | |
| } | |
| /* we ignore really small values | |
| * anyway, they will be 0 after shrinking | |
| * to q15_t | |
| */ | |
| base = base - 16; | |
| sum = 0; | |
| for (i = 0; i < dim_vec; i++) | |
| { | |
| if (vec_in[i] > base) | |
| { | |
| shift = (uint8_t)__USAT(vec_in[i] - base, 5); | |
| sum += 0x1 << shift; | |
| } | |
| } | |
| /* This is effectively (0x1 << 32) / sum */ | |
| int64_t div_base = 0x100000000LL; | |
| int output_base = (int32_t)(div_base / sum); | |
| /* Final confidence will be output_base >> ( 17 - (vec_in[i] - base) ) | |
| * so 32768 (0x1<<15) -> 100% confidence when sum = 0x1 << 16, output_base = 0x1 << 16 | |
| * and vec_in[i]-base = 16 | |
| */ | |
| for (i = 0; i < dim_vec; i++) | |
| { | |
| if (vec_in[i] > base) | |
| { | |
| /* Here minimum value of 17+base-vec[i] will be 1 */ | |
| shift = (uint8_t)__USAT(17 + base - vec_in[i], 5); | |
| p_out[i] = (q15_t)__SSAT((output_base >> shift), 16); | |
| } | |
| else | |
| { | |
| p_out[i] = 0; | |
| } | |
| } | |
| } | |
| /** | |
| * @} end of Softmax group | |
| */ | |