cera 0.5.3

Rust-native LLM inference engine
Documentation 
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
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169

#include <metal_stdlib>
#include <metal_math>
#include <metal_texture>
using namespace metal;

#line 56 "cera/src/backend/shaders/slang/rmsnorm.slang"
struct KernelContext_0
{
    packed_uint4 device* p_metal_0;
    float device* src_buf_0;
    float device* dst_buf_0;
    float device* w_metal_0;
    array<float, int(256)> threadgroup* scratch_0;
};


#line 50
float block_sum_0(uint tid_0, float v_0, KernelContext_0 thread* kernelContext_0)
{



    float sg_0 = simd_sum(v_0);
    if((tid_0 & 31U) == 0U)
    {

#line 56
        (*kernelContext_0->scratch_0)[tid_0 >> 5U] = sg_0;

#line 56
    }
    threadgroup_barrier(mem_flags::mem_threadgroup);

#line 57
    float lane_0;
    if(tid_0 < 8U)
    {

#line 58
        lane_0 = (*kernelContext_0->scratch_0)[tid_0];

#line 58
    }
    else
    {

#line 58
        lane_0 = 0.0f;

#line 58
    }
    float total_0 = simd_sum(lane_0);
    if(tid_0 == 0U)
    {

#line 60
        (*kernelContext_0->scratch_0)[int(0)] = total_0;

#line 60
    }
    threadgroup_barrier(mem_flags::mem_threadgroup);
    float _S1 = (*kernelContext_0->scratch_0)[int(0)];

#line 77
    return _S1;
}



[[kernel]] void rmsnorm(uint3 lid_0 [[thread_position_in_threadgroup]], packed_uint4 device* p_metal_1 [[buffer(3)]], float device* src_buf_1 [[buffer(0)]], float device* dst_buf_1 [[buffer(1)]], float device* w_metal_1 [[buffer(2)]])
{

#line 82
    thread KernelContext_0 kernelContext_1;

#line 82
    (&kernelContext_1)->p_metal_0 = p_metal_1;

#line 82
    (&kernelContext_1)->src_buf_0 = src_buf_1;

#line 82
    (&kernelContext_1)->dst_buf_0 = dst_buf_1;

#line 82
    (&kernelContext_1)->w_metal_0 = w_metal_1;

#line 82
    threadgroup array<float, int(256)> scratch_1;

#line 82
    (&kernelContext_1)->scratch_0 = &scratch_1;
    uint tid_1 = lid_0.x;

#line 89
    uint n_0 = (uint4(*(p_metal_1+int(0))) ).x;
    float eps_0 = (as_type<float>(((uint4(*(p_metal_1+int(0))) ).y)));

#line 90
    uint i_0 = tid_1;

#line 90
    float partial_0 = 0.0f;


    for(;;)
    {

#line 93
        if(i_0 < n_0)
        {
        }
        else
        {

#line 93
            break;
        }

#line 94
        float v_1 = (&kernelContext_1)->src_buf_0[i_0];
        float partial_1 = partial_0 + v_1 * v_1;

#line 93
        i_0 = i_0 + 256U;

#line 93
        partial_0 = partial_1;

#line 93
    }

#line 93
    float _S2 = block_sum_0(tid_1, partial_0, &kernelContext_1);



    float _S3 = 1.0f / sqrt(_S2 / float(n_0) + eps_0);

#line 97
    i_0 = tid_1;

    for(;;)
    {

#line 99
        if(i_0 < n_0)
        {
        }
        else
        {

#line 99
            break;
        }

#line 100
        *((&kernelContext_1)->dst_buf_0+i_0) = (&kernelContext_1)->src_buf_0[i_0] * _S3 * (&kernelContext_1)->w_metal_0[i_0];

#line 99
        i_0 = i_0 + 256U;

#line 99
    }

#line 123
    return;
}