arcweight 0.3.0

A high-performance, modular library for weighted finite state transducers with comprehensive examples and benchmarks
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
170
171
172
173
174
175
176
177

use arcweight::prelude::*;
use criterion::{criterion_group, criterion_main, Criterion};
#[cfg(feature = "parallel")]
use rayon::prelude::*;
use std::hint::black_box;
use std::time::Duration;

fn create_large_fst(n: usize) -> VectorFst<TropicalWeight> {
    let mut fst = VectorFst::new();
    let mut states = Vec::new();

    for _ in 0..=n {
        states.push(fst.add_state());
    }

    fst.set_start(states[0]);
    fst.set_final(states[n], TropicalWeight::one());

    for i in 0..n {
        for j in 1..=5 {
            fst.add_arc(
                states[i],
                Arc::new(
                    j as u32,
                    j as u32,
                    TropicalWeight::new(j as f32),
                    states[i + 1],
                ),
            );
        }
    }

    fst
}

pub fn bench_parallel_arc_processing(c: &mut Criterion) {
    let mut group = c.benchmark_group("parallel_arc_processing");
    group.measurement_time(Duration::from_secs(10));

    let fst = create_large_fst(1000);

    group.bench_function("sequential_arc_sum", |b| {
        b.iter(|| {
            let fst = black_box(&fst);
            let mut total = 0.0f32;
            for state in fst.states() {
                for arc in fst.arcs(state) {
                    total += arc.weight.value();
                }
            }
            black_box(total);
        })
    });

    #[cfg(feature = "parallel")]
    group.bench_function("parallel_arc_sum", |b| {
        b.iter(|| {
            let fst = black_box(&fst);
            let states: Vec<_> = fst.states().collect();
            let total: f32 = states
                .par_iter()
                .flat_map(|state| fst.arcs(*state).collect::<Vec<_>>())
                .map(|arc| *arc.weight.value())
                .sum();
            black_box(total);
        })
    });
    #[cfg(not(feature = "parallel"))]
    group.bench_function("parallel_arc_sum", |b| {
        b.iter(|| {
            let fst = black_box(&fst);
            let states: Vec<_> = fst.states().collect();
            let total: f32 = states
                .iter()
                .flat_map(|state| fst.arcs(*state).collect::<Vec<_>>())
                .map(|arc| *arc.weight.value())
                .sum();
            black_box(total);
        })
    });

    group.finish();
}

pub fn bench_parallel_state_processing(c: &mut Criterion) {
    let mut group = c.benchmark_group("parallel_state_processing");
    group.measurement_time(Duration::from_secs(10));

    let fst = create_large_fst(1000);

    group.bench_function("sequential_state_count", |b| {
        b.iter(|| {
            let fst = black_box(&fst);
            let count = fst.num_states();
            black_box(count);
        })
    });

    #[cfg(feature = "parallel")]
    group.bench_function("parallel_state_analysis", |b| {
        b.iter(|| {
            let fst = black_box(&fst);
            let states: Vec<_> = fst.states().collect();
            let analysis: Vec<(u32, usize)> = states
                .par_iter()
                .map(|&state| (state, fst.num_arcs(state)))
                .collect();
            black_box(analysis);
        })
    });
    #[cfg(not(feature = "parallel"))]
    group.bench_function("parallel_state_analysis", |b| {
        b.iter(|| {
            let fst = black_box(&fst);
            let states: Vec<_> = fst.states().collect();
            let analysis: Vec<(u32, usize)> = states
                .iter()
                .map(|&state| (state, fst.num_arcs(state)))
                .collect();
            black_box(analysis);
        })
    });

    group.finish();
}

pub fn bench_parallel_final_states(c: &mut Criterion) {
    let mut group = c.benchmark_group("parallel_final_states");
    group.measurement_time(Duration::from_secs(10));

    let fst = create_large_fst(1000);

    group.bench_function("sequential_final_check", |b| {
        b.iter(|| {
            let fst = black_box(&fst);
            let mut final_count = 0usize;
            for state in fst.states() {
                if fst.is_final(state) {
                    final_count += 1;
                }
            }
            black_box(final_count);
        })
    });

    #[cfg(feature = "parallel")]
    group.bench_function("parallel_final_check", |b| {
        b.iter(|| {
            let fst = black_box(&fst);
            let states: Vec<_> = fst.states().collect();
            let final_count = states
                .par_iter()
                .filter(|&&state| fst.is_final(state))
                .count();
            black_box(final_count);
        })
    });
    #[cfg(not(feature = "parallel"))]
    group.bench_function("parallel_final_check", |b| {
        b.iter(|| {
            let fst = black_box(&fst);
            let states: Vec<_> = fst.states().collect();
            let final_count = states.iter().filter(|&&state| fst.is_final(state)).count();
            black_box(final_count);
        })
    });

    group.finish();
}

criterion_group!(
    benches,
    bench_parallel_arc_processing,
    bench_parallel_state_processing,
    bench_parallel_final_states
);
criterion_main!(benches);