Gitiles
Code Review Sign In
review.mlplatform.org / ml / ComputeLibrary / 1009e87127e08a58561373c3f83f22b5f0922641 / . / arm_compute / runtime / CL / functions / CLLSTMLayer.h
blob: 1a8b33463d26c0b824938cdcf85b592701d622dd [file] [log] [blame]
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]1/*
Michele Di Giorgiod9eaf612020-07-08 11:12:57 +0100[diff] [blame]2 * Copyright (c) 2018-2020 Arm Limited.
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]3 *
4 * SPDX-License-Identifier: MIT
5 *
6 * Permission is hereby granted, free of charge, to any person obtaining a copy
7 * of this software and associated documentation files (the "Software"), to
8 * deal in the Software without restriction, including without limitation the
9 * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
10 * sell copies of the Software, and to permit persons to whom the Software is
11 * furnished to do so, subject to the following conditions:
12 *
13 * The above copyright notice and this permission notice shall be included in all
14 * copies or substantial portions of the Software.
15 *
16 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
17 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
18 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
19 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
20 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
21 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
22 * SOFTWARE.
23 */
Michalis Spyrouf4643372019-11-29 16:17:13 +0000[diff] [blame]24#ifndef ARM_COMPUTE_CLLSTMLAYER_H
25#define ARM_COMPUTE_CLLSTMLAYER_H
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]26
27#include "arm_compute/runtime/IFunction.h"
28
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]29#include "arm_compute/core/CL/kernels/CLCopyKernel.h"
Georgios Pinitasdbfc2dc2019-04-02 12:51:21 +0100[diff] [blame]30#include "arm_compute/core/CL/kernels/CLMemsetKernel.h"
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]31#include "arm_compute/core/Types.h"
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]32#include "arm_compute/runtime/CL/CLTensor.h"
Georgios Pinitasab23dd02020-07-06 14:57:36 +0100[diff] [blame]33#include "arm_compute/runtime/CL/functions/CLActivationLayer.h"
Georgios Pinitas09f24972019-05-17 18:14:40 +0100[diff] [blame]34#include "arm_compute/runtime/CL/functions/CLConcatenateLayer.h"
giuros01164a2722018-11-20 18:34:46 +0000[diff] [blame]35#include "arm_compute/runtime/CL/functions/CLElementwiseOperations.h"
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]36#include "arm_compute/runtime/CL/functions/CLFullyConnectedLayer.h"
37#include "arm_compute/runtime/CL/functions/CLGEMM.h"
Michele Di Giorgio39438b42019-06-04 12:41:45 +0100[diff] [blame]38#include "arm_compute/runtime/CL/functions/CLMeanStdDevNormalizationLayer.h"
Michalis Spyrou1009e872020-07-27 12:48:34 +0100[diff] [blame^]39#include "arm_compute/runtime/CL/functions/CLPixelWiseMultiplication.h"
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]40#include "arm_compute/runtime/IMemoryManager.h"
Georgios Pinitas26014cf2019-09-09 19:00:57 +0100[diff] [blame]41#include "arm_compute/runtime/MemoryGroup.h"
Michalis Spyrou25f45a42018-08-08 12:53:05 +0100[diff] [blame]42#include "arm_compute/runtime/common/LSTMParams.h"
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]43
44#include <memory>
45
46namespace arm_compute
47{
48class ICLTensor;
49
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]50/** This function performs a single time step in a Long Short-Term Memory (LSTM) layer.
51 *
52 */
53class CLLSTMLayer : public IFunction
54{
55public:
56 /** Default constructor */
57 CLLSTMLayer(std::shared_ptr<IMemoryManager> memory_manager = nullptr);
58 /** Initialize function's tensors.
59 *
Georgios Pinitas8bc745d2018-07-18 19:51:24 +0100[diff] [blame]60 * @param[in] input Source tensor. Input is a 2D tensor with dimensions [input_size, batch_size]. Data types supported: F16/F32.
61 * @param[in] input_to_forget_weights 2D weights tensor with dimensions [input_size, num_units]. Data type supported: Same as @p input.
62 * @param[in] input_to_cell_weights 2D weights tensor with dimensions [input_size, num_units]. Data type supported: Same as @p input.
63 * @param[in] input_to_output_weights 2D weights tensor with dimensions [input_size, num_units]. Data type supported: Same as @p input.
64 * @param[in] recurrent_to_forget_weights 2D weights tensor with dimensions [output_size, num_units]. Data type supported: Same as @p input.
65 * @param[in] recurrent_to_cell_weights 2D weights tensor with dimensions [output_size, num_units]. Data type supported: Same as @p input.
66 * @param[in] recurrent_to_output_weights 2D weights tensor with dimensions [output_size, num_units]. Data type supported: Same as @p input.
67 * @param[in] forget_gate_bias 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
68 * @param[in] cell_bias 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
69 * @param[in] output_gate_bias 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
70 * @param[in] output_state_in 2D weights tensor with dimensions [output_size, batch_size]. Data type supported: Same as @p input.
71 * @param[in] cell_state_in 2D tensor with dimensions [num_units, batch_size]. Data type supported: Same as @p input.
72 * @param[out] scratch_buffer 2D tensor with dimensions [num_units * 4, batch_size] with CIFG or [num_units * 3, batch_size] without CIGF. Data type supported: Same as @p input.
73 * @param[out] output_state_out 2D weights tensor with dimensions [output_size, batch_size]. Data type supported: Same as @p input.
74 * @param[out] cell_state_out 2D tensor with dimensions [num_units, batch_size]. Data type supported: Same as @p input.
75 * @param[out] output Destination tensor. Output is a 2D tensor with dimensions [output_size, batch_size].
76 * Data types supported: Same as @p input.
Michele Di Giorgio25d97752020-03-04 18:08:47 +0000[diff] [blame]77 * @param[in] lstm_params Weights tensors used in peephole optimization:
78 * input_to_input_weights 2D weights tensor with dimensions [input_size, num_units]. Data type supported: Same as @p input.
79 * recurrent_to_input_weights 2D weights tensor with dimensions [output_size, num_units]. Data type supported: Same as @p input.
80 * cell_to_input_weights 1D weights tensor with dimensions [num_units]. Can be nullptr. Data type supported: Same as @p input.
81 * cell_to_forget_weights 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
82 * cell_to_output_weights 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
83 * input_gate_bias 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input
84 * projection_weights 2D weights tensor with dimensions [output_size, num_units]. Data type supported: Same as @p input.
85 * projection_bias 1D weights tensor with dimensions [output_size]. Data type supported: Same as @p input.
86 * input_layer_norm_weights 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
87 * forget_layer_norm_weights 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
88 * cell_layer_norm_weights 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
89 * output_layer_norm_weights 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
Georgios Pinitas8bc745d2018-07-18 19:51:24 +0100[diff] [blame]90 * @param[in] activation_info Contains activation information described in @ref ActivationLayerInfo.
Michele Di Giorgio25d97752020-03-04 18:08:47 +0000[diff] [blame]91 * @param[in] cell_threshold (Optional) The clipping threshold for the cell state, such that values are bound within [-cell_clip, cell_clip].
92 * If set to 0.0f then clipping is disabled.
93 * @param[in] projection_threshold (Optional) The clipping threshold for the output from the projection layer, such that values are bound within [-proj_clip, proj_clip].
Michele Di Giorgio39438b42019-06-04 12:41:45 +0100[diff] [blame]94 * If set to 0.0f then clipping is disabled.
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]95 */
Georgios Pinitas8bc745d2018-07-18 19:51:24 +0100[diff] [blame]96 void configure(const ICLTensor *input,
97 const ICLTensor *input_to_forget_weights, const ICLTensor *input_to_cell_weights, const ICLTensor *input_to_output_weights,
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]98 const ICLTensor *recurrent_to_forget_weights, const ICLTensor *recurrent_to_cell_weights, const ICLTensor *recurrent_to_output_weights,
Georgios Pinitas8bc745d2018-07-18 19:51:24 +0100[diff] [blame]99 const ICLTensor *forget_gate_bias, const ICLTensor *cell_bias, const ICLTensor *output_gate_bias,
Michalis Spyrou1009e872020-07-27 12:48:34 +0100[diff] [blame^]100 const ICLTensor *output_state_in, ICLTensor *cell_state_in,
Georgios Pinitas8bc745d2018-07-18 19:51:24 +0100[diff] [blame]101 ICLTensor *scratch_buffer, ICLTensor *output_state_out, ICLTensor *cell_state_out, ICLTensor *output,
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]102 const LSTMParams<ICLTensor> &lstm_params, const ActivationLayerInfo &activation_info, float cell_threshold = 0.f, float projection_threshold = 0.f);
Manuel Bottini2b84be52020-04-08 10:15:51 +0100[diff] [blame]103 /** Initialize function's tensors.
104 *
105 * @param[in] compile_context The compile context to be used.
106 * @param[in] input Source tensor. Input is a 2D tensor with dimensions [input_size, batch_size]. Data types supported: F16/F32.
107 * @param[in] input_to_forget_weights 2D weights tensor with dimensions [input_size, num_units]. Data type supported: Same as @p input.
108 * @param[in] input_to_cell_weights 2D weights tensor with dimensions [input_size, num_units]. Data type supported: Same as @p input.
109 * @param[in] input_to_output_weights 2D weights tensor with dimensions [input_size, num_units]. Data type supported: Same as @p input.
110 * @param[in] recurrent_to_forget_weights 2D weights tensor with dimensions [output_size, num_units]. Data type supported: Same as @p input.
111 * @param[in] recurrent_to_cell_weights 2D weights tensor with dimensions [output_size, num_units]. Data type supported: Same as @p input.
112 * @param[in] recurrent_to_output_weights 2D weights tensor with dimensions [output_size, num_units]. Data type supported: Same as @p input.
113 * @param[in] forget_gate_bias 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
114 * @param[in] cell_bias 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
115 * @param[in] output_gate_bias 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
116 * @param[in] output_state_in 2D weights tensor with dimensions [output_size, batch_size]. Data type supported: Same as @p input.
117 * @param[in] cell_state_in 2D tensor with dimensions [num_units, batch_size]. Data type supported: Same as @p input.
118 * @param[out] scratch_buffer 2D tensor with dimensions [num_units * 4, batch_size] with CIFG or [num_units * 3, batch_size] without CIGF. Data type supported: Same as @p input.
119 * @param[out] output_state_out 2D weights tensor with dimensions [output_size, batch_size]. Data type supported: Same as @p input.
120 * @param[out] cell_state_out 2D tensor with dimensions [num_units, batch_size]. Data type supported: Same as @p input.
121 * @param[out] output Destination tensor. Output is a 2D tensor with dimensions [output_size, batch_size].
122 * Data types supported: Same as @p input.
123 * @param[in] lstm_params Weights tensors used in peephole optimization:
124 * input_to_input_weights 2D weights tensor with dimensions [input_size, num_units]. Data type supported: Same as @p input.
125 * recurrent_to_input_weights 2D weights tensor with dimensions [output_size, num_units]. Data type supported: Same as @p input.
126 * cell_to_input_weights 1D weights tensor with dimensions [num_units]. Can be nullptr. Data type supported: Same as @p input.
127 * cell_to_forget_weights 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
128 * cell_to_output_weights 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
129 * input_gate_bias 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input
130 * projection_weights 2D weights tensor with dimensions [output_size, num_units]. Data type supported: Same as @p input.
131 * projection_bias 1D weights tensor with dimensions [output_size]. Data type supported: Same as @p input.
132 * input_layer_norm_weights 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
133 * forget_layer_norm_weights 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
134 * cell_layer_norm_weights 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
135 * output_layer_norm_weights 1D weights tensor with dimensions [num_units]. Data type supported: Same as @p input.
136 * @param[in] activation_info Contains activation information described in @ref ActivationLayerInfo.
137 * @param[in] cell_threshold (Optional) The clipping threshold for the cell state, such that values are bound within [-cell_clip, cell_clip].
138 * If set to 0.0f then clipping is disabled.
139 * @param[in] projection_threshold (Optional) The clipping threshold for the output from the projection layer, such that values are bound within [-proj_clip, proj_clip].
140 * If set to 0.0f then clipping is disabled.
141 */
142 void configure(const CLCompileContext &compile_context, const ICLTensor *input,
143 const ICLTensor *input_to_forget_weights, const ICLTensor *input_to_cell_weights, const ICLTensor *input_to_output_weights,
144 const ICLTensor *recurrent_to_forget_weights, const ICLTensor *recurrent_to_cell_weights, const ICLTensor *recurrent_to_output_weights,
145 const ICLTensor *forget_gate_bias, const ICLTensor *cell_bias, const ICLTensor *output_gate_bias,
Michalis Spyrou1009e872020-07-27 12:48:34 +0100[diff] [blame^]146 const ICLTensor *output_state_in, ICLTensor *cell_state_in,
Manuel Bottini2b84be52020-04-08 10:15:51 +0100[diff] [blame]147 ICLTensor *scratch_buffer, ICLTensor *output_state_out, ICLTensor *cell_state_out, ICLTensor *output,
148 const LSTMParams<ICLTensor> &lstm_params, const ActivationLayerInfo &activation_info, float cell_threshold = 0.f, float projection_threshold = 0.f);
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]149
150 /** Static function to check if given info will lead to a valid configuration of @ref CLLSTMLayer
151 *
Michele Di Giorgio39438b42019-06-04 12:41:45 +0100[diff] [blame]152 * @param[in] input Source tensor info. Input is a 2D tensor with dimensions [input_size, batch_size]. Data types supported: F16/F32.
153 * @param[in] input_to_forget_weights 2D weights tensor info with dimensions [input_size, num_units]. Data type supported: Same as @p input.
154 * @param[in] input_to_cell_weights 2D weights tensor info with dimensions [input_size, num_units]. Data type supported: Same as @p input.
155 * @param[in] input_to_output_weights 2D weights tensor info with dimensions [input_size, num_units]. Data type supported: Same as @p input.
156 * @param[in] recurrent_to_forget_weights 2D weights tensor info with dimensions [output_size, num_units]. Data type supported: Same as @p input.
157 * @param[in] recurrent_to_cell_weights 2D weights tensor info with dimensions [output_size, num_units]. Data type supported: Same as @p input.
158 * @param[in] recurrent_to_output_weights 2D weights tensor info with dimensions [output_size, num_units]. Data type supported: Same as @p input.
159 * @param[in] forget_gate_bias 1D weights tensor info with dimensions [num_units]. Data type supported: Same as @p input.
160 * @param[in] cell_bias 1D weights tensor info with dimensions [num_units]. Data type supported: Same as @p input.
161 * @param[in] output_gate_bias 1D weights tensor info with dimensions [num_units]. Data type supported: Same as @p input.
162 * @param[in] output_state_in 2D weights tensor info with dimensions [output_size, batch_size]. Data type supported: Same as @p input.
163 * @param[in] cell_state_in 2D tensor info with dimensions [num_units, batch_size]. Data type supported: Same as @p input.
164 * @param[in] scratch_buffer 2D tensor info with dimensions [num_units * 4, batch_size] with CIFG or [num_units * 3, batch_size] without CIGF.
165 * Data type supported: Same as @p input.
166 * @param[in] output_state_out 2D weights tensor info with dimensions [output_size, batch_size]. Data type supported: Same as @p input.
167 * @param[in] cell_state_out 2D tensor info with dimensions [num_units, batch_size]. Data type supported: Same as @p input.
168 * @param[in] output Destination tensor info. Output is a 2D tensor with dimensions [output_size, batch_size]. Data types supported: Same as @p input.
Michele Di Giorgio25d97752020-03-04 18:08:47 +0000[diff] [blame]169 * @param[in] lstm_params Weights tensors info used in peephole optimization:
170 * input_to_input_weights 2D weights tensor info with dimensions [input_size, num_units]. Data type supported: Same as @p input.
171 * recurrent_to_input_weights 2D weights tensor info with dimensions [output_size, num_units]. Data type supported: Same as @p input.
172 * cell_to_input_weights 1D weights tensor info with dimensions [num_units]. Can be nullptr. Data type supported: Same as @p input.
173 * cell_to_forget_weights 1D weights tensor info with dimensions [num_units]. Data type supported: Same as @p input.
174 * cell_to_output_weights 1D weights tensor info with dimensions [num_units]. Data type supported: Same as @p input.
175 * input_gate_bias 1D weights tensor info with dimensions [num_units]. Data type supported: Same as @p input
176 * projection_weights 2D weights tensor info with dimensions [output_size, num_units]. Data type supported: Same as @p input.
177 * projection_bias 1D weights tensor info with dimensions [output_size]. Data type supported: Same as @p input.
178 * input_layer_norm_weights 1D weights tensor info with dimensions [num_units]. Data type supported: Same as @p input.
179 * forget_layer_norm_weights 1D weights tensor info with dimensions [num_units]. Data type supported: Same as @p input.
180 * cell_layer_norm_weights 1D weights tensor info with dimensions [num_units]. Data type supported: Same as @p input.
181 * output_layer_norm_weights 1D weights tensor info with dimensions [num_units]. Data type supported: Same as @p input.
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]182 * @param[in] activation_info Contains activation information described in @ref ActivationLayerInfo.
Michele Di Giorgio25d97752020-03-04 18:08:47 +0000[diff] [blame]183 * @param[in] cell_threshold (Optional) The clipping threshold for the cell state, such that values are bound within [-cell_clip, cell_clip].
184 * If set to 0.0f then clipping is disabled.
185 * @param[in] projection_threshold (Optional) The clipping threshold for the output from the projection layer, such that values are bound within [-proj_clip, proj_clip].
Michele Di Giorgio39438b42019-06-04 12:41:45 +0100[diff] [blame]186 * If set to 0.0f then clipping is disabled.
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]187 *
188 * @return a status
189 */
Georgios Pinitas8bc745d2018-07-18 19:51:24 +0100[diff] [blame]190 static Status validate(const ITensorInfo *input,
191 const ITensorInfo *input_to_forget_weights, const ITensorInfo *input_to_cell_weights, const ITensorInfo *input_to_output_weights,
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]192 const ITensorInfo *recurrent_to_forget_weights, const ITensorInfo *recurrent_to_cell_weights, const ITensorInfo *recurrent_to_output_weights,
193 const ITensorInfo *forget_gate_bias, const ITensorInfo *cell_bias, const ITensorInfo *output_gate_bias,
Georgios Pinitas8bc745d2018-07-18 19:51:24 +0100[diff] [blame]194 const ITensorInfo *output_state_in, const ITensorInfo *cell_state_in,
195 const ITensorInfo *scratch_buffer, const ITensorInfo *output_state_out, const ITensorInfo *cell_state_out, const ITensorInfo *output,
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]196 const LSTMParams<ITensorInfo> &lstm_params, const ActivationLayerInfo &activation_info, float cell_threshold = 0.f, float projection_threshold = 0.f);
197
198 // Inherited methods overridden:
199 void run() override;
John Kesapidescafec8f2019-02-19 15:53:59 +0000[diff] [blame]200 void prepare() override;
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]201
202private:
Michalis Spyrou1009e872020-07-27 12:48:34 +0100[diff] [blame^]203 MemoryGroup _memory_group;
204 CLFullyConnectedLayer _fully_connected_input_gate;
205 CLArithmeticAddition _accum_input_gate1;
206 CLArithmeticSubtraction _subtract_input_gate;
207 CLPixelWiseMultiplication _pixelwise_mul_input_gate;
208 CLActivationLayer _activation_input_gate;
209 CLFullyConnectedLayer _fully_connected_forget_gate;
210 CLArithmeticAddition _accum_forget_gate1;
211 CLPixelWiseMultiplication _pixelwise_mul_forget_gate;
212 CLActivationLayer _activation_forget_gate;
213 CLFullyConnectedLayer _fully_connected_cell_state;
214 CLGEMM _gemm_cell_state1;
215 CLTransposeKernel _transpose_cell_state;
216 CLArithmeticAddition _accum_cell_state1;
217 CLArithmeticAddition _accum_cell_state2;
218 CLPixelWiseMultiplication _pixelwise_mul_cell_state1;
219 CLActivationLayer _activation_cell_state;
220 CLActivationLayer _cell_clip;
221 CLPixelWiseMultiplication _pixelwise_mul_cell_state2;
222 CLFullyConnectedLayer _fully_connected_output;
223 CLPixelWiseMultiplication _pixelwise_mul_output_state1;
224 CLArithmeticAddition _accum_output1;
225 CLActivationLayer _activation_output;
226 CLActivationLayer _activation_output_state;
227 CLPixelWiseMultiplication _pixelwise_mul_output_state2;
228 CLFullyConnectedLayer _fully_connected_output_state;
229 CLActivationLayer _projection_clip;
230 CLCopyKernel _copy_cell_state;
231 CLCopyKernel _copy_output;
232 CLConcatenateLayer _concat_scratch_buffer;
233 CLConcatenateLayer _concat_inputs_forget_gate;
234 CLConcatenateLayer _concat_weights_forget_gate;
235 CLConcatenateLayer _concat_weights_input_gate;
236 CLConcatenateLayer _concat_weights_output;
237 CLMemsetKernel _ones_memset_kernel;
238 CLMeanStdDevNormalizationLayer _mean_std_norm_input_gate;
239 CLPixelWiseMultiplication _pixelwise_mul_input_gate_coeff;
240 CLArithmeticAddition _accum_input_gate_bias;
241 CLMeanStdDevNormalizationLayer _mean_std_norm_forget_gate;
242 CLPixelWiseMultiplication _pixelwise_mul_forget_gate_coeff;
243 CLArithmeticAddition _accum_forget_gate_bias;
244 CLMeanStdDevNormalizationLayer _mean_std_norm_cell_gate;
245 CLPixelWiseMultiplication _pixelwise_mul_cell_gate_coeff;
246 CLArithmeticAddition _accum_cell_gate_bias;
247 CLMeanStdDevNormalizationLayer _mean_std_norm_output_gate;
248 CLPixelWiseMultiplication _pixelwise_mul_output_gate_coeff;
249 CLArithmeticAddition _accum_output_gate_bias;
250 CLTensor _input_gate_out1;
251 CLTensor _input_gate_out2;
252 CLTensor _input_gate_out3;
253 CLTensor _input_gate_out4;
254 CLTensor _forget_gate_out1;
255 CLTensor _forget_gate_out2;
256 CLTensor _forget_gate_out3;
257 CLTensor _forget_gate_out4;
258 CLTensor _forget_gate_out5;
259 CLTensor _forget_gate_out6;
260 CLTensor _cell_state_out1;
261 CLTensor _cell_state_out2;
262 CLTensor _cell_state_out3;
263 CLTensor _cell_state_out4;
264 CLTensor _cell_state_out5;
265 CLTensor _output1;
266 CLTensor _output2;
267 CLTensor _output3;
268 CLTensor _output4;
269 CLTensor _cell_state_activation;
270 CLTensor _output_state1;
271 CLTensor _ones;
272 CLTensor _input_layer_norm_out1;
273 CLTensor _input_layer_norm_out2;
274 CLTensor _forget_layer_norm_out1;
275 CLTensor _forget_layer_norm_out2;
276 CLTensor _cell_layer_norm_out1;
277 CLTensor _cell_layer_norm_out2;
278 CLTensor _output_layer_norm_out1;
279 CLTensor _output_layer_norm_out2;
280 bool _run_peephole_opt;
281 bool _run_cifg_opt;
282 bool _perform_cell_clipping;
283 bool _has_projection_weights;
284 bool _perform_projection_clipping;
285 bool _is_prepared;
286 bool _is_layer_norm_lstm;
Michalis Spyroubcedf512018-03-22 14:55:08 +0000[diff] [blame]287};
John Kesapidescafec8f2019-02-19 15:53:59 +0000[diff] [blame]288} // namespace arm_compute
Michalis Spyrouf4643372019-11-29 16:17:13 +0000[diff] [blame]289#endif /* ARM_COMPUTE_CLLSTMLAYER_H */