1 /*
2 * SPDX-FileCopyrightText: Copyright 2010-2024 Arm Limited and/or its affiliates <open-source-office@arm.com>
3 *
4 * SPDX-License-Identifier: Apache-2.0
5 *
6 * Licensed under the Apache License, Version 2.0 (the License); you may
7 * not use this file except in compliance with the License.
8 * You may obtain a copy of the License at
9 *
10 * www.apache.org/licenses/LICENSE-2.0
11 *
12 * Unless required by applicable law or agreed to in writing, software
13 * distributed under the License is distributed on an AS IS BASIS, WITHOUT
14 * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
15 * See the License for the specific language governing permissions and
16 * limitations under the License.
17 */
18
19 /* ----------------------------------------------------------------------
20 * Project: CMSIS NN Library
21 * Title: arm_fully_connected_per_channel_s8
22 * Description: Fully connected function compatible with TF Lite.
23 *
24 * $Date: 15 Aug 2024
25 * $Revision: V.1.0.0
26 *
27 * Target : Arm(R) M-Profile Architecture
28 *
29 * -------------------------------------------------------------------- */
30
31 #include "arm_nnfunctions.h"
32 #include "arm_nnsupportfunctions.h"
33
34 /**
35 * @ingroup Public
36 */
37
38 /**
39 * @addtogroup FC
40 * @{
41 */
42
43 /*
44 * S8 basic fully-connected and matrix multiplication layer function using per-channel quantization for TensorFlow Lite
45 *
46 * Refer header file for details.
47 *
48 */
arm_fully_connected_per_channel_s8(const cmsis_nn_context * ctx,const cmsis_nn_fc_params * fc_params,const cmsis_nn_per_channel_quant_params * quant_params,const cmsis_nn_dims * input_dims,const int8_t * input_data,const cmsis_nn_dims * filter_dims,const int8_t * kernel,const cmsis_nn_dims * bias_dims,const int32_t * bias_data,const cmsis_nn_dims * output_dims,int8_t * output_data)49 arm_cmsis_nn_status arm_fully_connected_per_channel_s8(const cmsis_nn_context *ctx,
50 const cmsis_nn_fc_params *fc_params,
51 const cmsis_nn_per_channel_quant_params *quant_params,
52 const cmsis_nn_dims *input_dims,
53 const int8_t *input_data,
54 const cmsis_nn_dims *filter_dims,
55 const int8_t *kernel,
56 const cmsis_nn_dims *bias_dims,
57 const int32_t *bias_data,
58 const cmsis_nn_dims *output_dims,
59 int8_t *output_data)
60 {
61 (void)bias_dims;
62
63 int32_t batch_cnt = input_dims->n;
64
65 #if defined(ARM_MATH_MVEI)
66 if (ctx->buf == NULL)
67 {
68 return (ARM_CMSIS_NN_ARG_ERROR);
69 }
70 #endif
71
72 const int32_t *kernel_sum = (const int32_t *)ctx->buf;
73
74 while (batch_cnt)
75 {
76
77 arm_nn_vec_mat_mult_t_per_ch_s8(input_data,
78 kernel,
79 kernel_sum,
80 bias_data,
81 output_data,
82 fc_params->input_offset,
83 fc_params->output_offset,
84 quant_params->multiplier,
85 quant_params->shift,
86 filter_dims->n, /* col_dim or accum_depth */
87 output_dims->c, /* row_dim or output_depth */
88 fc_params->activation.min,
89 fc_params->activation.max,
90 1L,
91 fc_params->filter_offset);
92
93 input_data += filter_dims->n;
94 output_data += output_dims->c;
95 batch_cnt--;
96 }
97 return (ARM_CMSIS_NN_SUCCESS);
98 }
99
100 /**
101 * @} end of FC group
102 */
103
