1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
|
/*******************************************************************************
* Copyright 2018 Intel Corporation
*
* 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
*
* http://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.
*******************************************************************************/
/*
* Cell execution LSTM
*/
#include "math_utils.hpp"
#include "mkldnn_thread.hpp"
#include "../simple_q10n.hpp"
#include "ref_rnn.hpp"
namespace mkldnn {
namespace impl {
namespace cpu {
using namespace mkldnn::impl::utils;
using namespace mkldnn::impl::math;
using namespace rnn_utils;
template <>
rnn_elemwise_sig(ref_rnn_fwd_f32_t::lstm_elemwise) {
ws_gates_aoc_t ws_gates(rnn, ws_gates_);
bias_aoc_t bias(rnn, bias_);
ws_states_aoc_t states_t_l(rnn, states_t_l_);
ws_states_aoc_t c_states_t_l(rnn, c_states_t_l_);
ws_states_aoc_t c_states_tm1_l(rnn, c_states_tm1_l_);
parallel_nd(rnn.mb, [&](int i) {
PRAGMA_OMP_SIMD()
for (int j = 0; j < rnn.dic; j++) {
ws_gates(i, 0, j) = logistic_fwd(ws_gates(i, 0, j) + bias(0, j));
ws_gates(i, 1, j) = logistic_fwd(ws_gates(i, 1, j) + bias(1, j));
ws_gates(i, 2, j) = tanh_fwd(ws_gates(i, 2, j) + bias(2, j));
ws_gates(i, 3, j) = logistic_fwd(ws_gates(i, 3, j) + bias(3, j));
float tmp = ws_gates(i, 1, j) * c_states_tm1_l(i, j)
+ ws_gates(i, 0, j) * ws_gates(i, 2, j);
states_t_l(i, j) = ws_gates(i, 3, j) * tanh_fwd(tmp);
c_states_t_l(i, j) = tmp;
}
});
}
template <>
rnn_elemwise_sig(ref_rnn_fwd_u8s8_t::lstm_elemwise) {
ws_gates_aoc_s32_t ws_gates_s32(rnn, ws_gates_);
bias_aoc_t bias(rnn, bias_);
ws_states_aoc_u8_t states_t_l(rnn, states_t_l_);
ws_states_aoc_t c_states_t_l(rnn, c_states_t_l_);
ws_states_aoc_t c_states_tm1_l(rnn, c_states_tm1_l_);
float *weights_scales = pd()->attr()->rnn_weights_qparams_.scales_;
float data_shift = pd()->attr()->rnn_data_qparams_.shift_;
float data_scale = pd()->attr()->rnn_data_qparams_.scale_;
auto q_d = [&](float f) {
float qf = f * data_scale + data_shift;
return qz_a1b0<float, src_data_t>()(qf);
};
auto deq_w = [&](acc_data_t s, int gate, int j) {
return pd()->attr()->rnn_weights_qparams_.mask_ == 0 ?
saturate<float>(s) * (1.f / (weights_scales[0] * data_scale)) :
saturate<float>(s) * (1.f / (weights_scales[gate * rnn.dic + j]
* data_scale));
};
parallel_nd(rnn.mb, [&](int i) {
PRAGMA_OMP_SIMD()
for (int j = 0; j < rnn.dic; j++) {
float G0 = logistic_fwd<float>(
deq_w(ws_gates_s32(i, 0, j), 0, j) + bias(0, j));
float G1 = logistic_fwd<float>(
deq_w(ws_gates_s32(i, 1, j), 1, j) + bias(1, j));
float G2 = tanh_fwd<float>(
deq_w(ws_gates_s32(i, 2, j), 2, j) + bias(2, j));
float G3 = logistic_fwd<float>(
deq_w(ws_gates_s32(i, 3, j), 3, j) + bias(3, j));
float tmp = G1 * c_states_tm1_l(i, j) + G0 * G2;
states_t_l(i, j) = q_d(G3 * tanh_fwd(tmp));
c_states_t_l(i, j) = tmp;
}
});
}
template <>
rnn_elemwise_sig(ref_rnn_bwd_f32_t::lstm_elemwise) {
ws_gates_aoc_t ws_gates(rnn, ws_gates_);
bias_aoc_t bias(rnn, bias_);
ws_states_aoc_t c_states_t_l(rnn, c_states_t_l_);
ws_states_aoc_t c_states_tm1_l(rnn, c_states_tm1_l_);
ws_diff_states_aoc_t diff_states_t_l(rnn, diff_states_t_l_);
ws_diff_states_aoc_t diff_states_tp1_l(rnn, diff_states_tp1_l_);
ws_diff_states_aoc_t diff_states_t_lp1(rnn, diff_states_t_lp1_);
parallel_nd(rnn.mb, [&](int i) {
PRAGMA_OMP_SIMD()
for (int j = 0; j < rnn.dic; j++) {
float Ct = c_states_t_l(i, j);
/// @todo save it in the workspace in fwd pass or recompute it to
/// save bw
float tanhCt = tanh_fwd(Ct);
// we have 2 incoming diffs on Ht
float dHt = diff_states_tp1_l(0, i, j)
+ diff_states_t_lp1(rnn.n_states, i, j);
float dCt = diff_states_tp1_l(1, i, j)
+ one_m_square(tanhCt) * ws_gates(i, 3, j) * dHt;
float dG1 = c_states_tm1_l(i, j) * dCt
* x_m_square(ws_gates(i, 1, j));
float dG0 = ws_gates(i, 2, j) * dCt * x_m_square(ws_gates(i, 0, j));
float dG3 = tanhCt * dHt * x_m_square(ws_gates(i, 3, j));
float dG2
= ws_gates(i, 0, j) * dCt * one_m_square(ws_gates(i, 2, j));
diff_states_t_l(1, i, j) = dCt * ws_gates(i, 1, j);
ws_gates(i, 0, j) = dG0;
ws_gates(i, 1, j) = dG1;
ws_gates(i, 2, j) = dG2;
ws_gates(i, 3, j) = dG3;
}
});
}
}
}
}
|