fasttext 0.8.0

fastText pure Rust implementation
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

// SSE2 SIMD kernels (x86_64 baseline).

use std::arch::x86_64::*;

use crate::matrix::{DenseMatrix, Matrix};
use crate::vector::Vector;

#[inline]
pub(crate) fn dot(a: &[f32], b: &[f32]) -> f32 {
    debug_assert_eq!(a.len(), b.len());
    let n = a.len();
    let chunks = n / 16;
    let remainder = n % 16;

    unsafe {
        let mut sum0 = _mm_setzero_ps();
        let mut sum1 = _mm_setzero_ps();
        let mut sum2 = _mm_setzero_ps();
        let mut sum3 = _mm_setzero_ps();

        let a_ptr = a.as_ptr();
        let b_ptr = b.as_ptr();

        for i in 0..chunks {
            let offset = i * 16;
            let a0 = _mm_loadu_ps(a_ptr.add(offset));
            let a1 = _mm_loadu_ps(a_ptr.add(offset + 4));
            let a2 = _mm_loadu_ps(a_ptr.add(offset + 8));
            let a3 = _mm_loadu_ps(a_ptr.add(offset + 12));
            let b0 = _mm_loadu_ps(b_ptr.add(offset));
            let b1 = _mm_loadu_ps(b_ptr.add(offset + 4));
            let b2 = _mm_loadu_ps(b_ptr.add(offset + 8));
            let b3 = _mm_loadu_ps(b_ptr.add(offset + 12));
            sum0 = _mm_add_ps(sum0, _mm_mul_ps(a0, b0));
            sum1 = _mm_add_ps(sum1, _mm_mul_ps(a1, b1));
            sum2 = _mm_add_ps(sum2, _mm_mul_ps(a2, b2));
            sum3 = _mm_add_ps(sum3, _mm_mul_ps(a3, b3));
        }

        sum0 = _mm_add_ps(sum0, sum1);
        sum2 = _mm_add_ps(sum2, sum3);
        sum0 = _mm_add_ps(sum0, sum2);

        let hi = _mm_movehl_ps(sum0, sum0);
        let s = _mm_add_ps(sum0, hi);
        let s = _mm_add_ss(s, _mm_shuffle_ps(s, s, 1));
        let mut result = _mm_cvtss_f32(s);

        let rem_start = chunks * 16;
        for i in 0..remainder {
            result += a[rem_start + i] * b[rem_start + i];
        }

        result
    }
}

#[inline]
pub(crate) fn add_vector(dest: &mut [f32], src: &[f32], scale: f32) {
    debug_assert_eq!(dest.len(), src.len());
    let n = dest.len();
    let chunks = n / 16;
    let remainder = n % 16;

    unsafe {
        let scale_v = _mm_set1_ps(scale);
        let d_ptr = dest.as_mut_ptr();
        let s_ptr = src.as_ptr();

        for i in 0..chunks {
            let offset = i * 16;
            let d0 = _mm_loadu_ps(d_ptr.add(offset));
            let d1 = _mm_loadu_ps(d_ptr.add(offset + 4));
            let d2 = _mm_loadu_ps(d_ptr.add(offset + 8));
            let d3 = _mm_loadu_ps(d_ptr.add(offset + 12));
            let s0 = _mm_loadu_ps(s_ptr.add(offset));
            let s1 = _mm_loadu_ps(s_ptr.add(offset + 4));
            let s2 = _mm_loadu_ps(s_ptr.add(offset + 8));
            let s3 = _mm_loadu_ps(s_ptr.add(offset + 12));
            _mm_storeu_ps(d_ptr.add(offset), _mm_add_ps(d0, _mm_mul_ps(scale_v, s0)));
            _mm_storeu_ps(
                d_ptr.add(offset + 4),
                _mm_add_ps(d1, _mm_mul_ps(scale_v, s1)),
            );
            _mm_storeu_ps(
                d_ptr.add(offset + 8),
                _mm_add_ps(d2, _mm_mul_ps(scale_v, s2)),
            );
            _mm_storeu_ps(
                d_ptr.add(offset + 12),
                _mm_add_ps(d3, _mm_mul_ps(scale_v, s3)),
            );
        }

        let rem_start = chunks * 16;
        for i in 0..remainder {
            dest[rem_start + i] += scale * src[rem_start + i];
        }
    }
}

pub(crate) fn average_rows(x: &mut Vector, rows: &[i32], mat: &DenseMatrix) {
    if rows.is_empty() {
        x.zero();
        return;
    }

    let n = mat.cols() as usize;

    let first_row = mat.row(rows[0] as i64);
    x.data_mut().copy_from_slice(first_row);

    for &row_idx in &rows[1..] {
        let row = mat.row(row_idx as i64);
        add_vector(x.data_mut(), row, 1.0);
    }

    let scale = 1.0 / rows.len() as f32;
    unsafe {
        let scale_v = _mm_set1_ps(scale);
        let chunks = n / 4;
        let remainder = n % 4;
        let d_ptr = x.data_mut().as_mut_ptr();

        for i in 0..chunks {
            let offset = i * 4;
            let v = _mm_loadu_ps(d_ptr.add(offset));
            _mm_storeu_ps(d_ptr.add(offset), _mm_mul_ps(v, scale_v));
        }

        let rem_start = chunks * 4;
        for i in 0..remainder {
            x.data_mut()[rem_start + i] *= scale;
        }
    }
}