boostr 0.1.0

ML framework built on numr - attention, quantization, model architectures
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

use super::*;
use numr::runtime::cpu::CpuRuntime;

struct MockModel {
    vocab_size: usize,
    name: String,
    fixed_logits: Option<Vec<f32>>,
}

impl MockModel {
    fn new(vocab_size: usize, name: &str) -> Self {
        Self {
            vocab_size,
            name: name.to_string(),
            fixed_logits: None,
        }
    }

    fn with_fixed_logits(mut self, logits: Vec<f32>) -> Self {
        self.fixed_logits = Some(logits);
        self
    }
}

impl SpeculativeModel<CpuRuntime> for MockModel {
    fn forward(&mut self, _input_tokens: &[TokenId], _position: usize) -> Result<Vec<f32>> {
        if let Some(ref logits) = self.fixed_logits {
            Ok(logits.clone())
        } else {
            Ok(vec![0.0; self.vocab_size])
        }
    }

    fn vocab_size(&self) -> usize {
        self.vocab_size
    }

    fn reset(&mut self) -> Result<()> {
        Ok(())
    }

    fn name(&self) -> &str {
        &self.name
    }
}

#[test]
fn test_speculative_config_default() {
    let config = SpeculativeConfig::default();
    assert_eq!(config.num_speculative_tokens, 4);
    assert!((config.draft_temperature - 0.7).abs() < 1e-6);
}

#[test]
fn test_speculative_stats() {
    let stats = SpeculativeStats {
        accepted_tokens: 80,
        rejected_tokens: 20,
        total_tokens: 100,
        target_forward_passes: 25,
        draft_forward_passes: 100,
        ..Default::default()
    };

    assert!((stats.acceptance_rate() - 0.8).abs() < 1e-6);
    assert!((stats.tokens_per_forward() - 4.0).abs() < 1e-6);
}

#[test]
fn test_speculative_executor_creation() {
    let draft = MockModel::new(100, "draft");
    let target = MockModel::new(100, "target");
    let config = SpeculativeConfig::default();

    let executor = SpeculativeExecutor::<CpuRuntime, _, _>::new(draft, target, config);

    assert_eq!(executor.current_depth(), 4);
    assert_eq!(executor.stats().total_tokens, 0);
}

#[test]
fn test_speculative_executor_generate() {
    let mut logits = vec![-10.0; 100];
    logits[5] = 0.0;
    logits[2] = -100.0;

    let draft = MockModel::new(100, "draft").with_fixed_logits(logits.clone());
    let target = MockModel::new(100, "target").with_fixed_logits(logits);

    let config = SpeculativeConfig {
        num_speculative_tokens: 2,
        seed: Some(12345),
        ..Default::default()
    };

    let mut executor = SpeculativeExecutor::<CpuRuntime, _, _>::new(draft, target, config);

    let prompt = vec![1, 2, 3];
    let result = executor.generate(&prompt, 5).unwrap();

    assert!(!result.is_empty());
    assert!(result.len() <= 5);

    let stats = executor.stats();
    assert!(stats.total_tokens > 0);
    assert!(stats.target_forward_passes > 0);
}

#[test]
fn test_verification_result() {
    let result = VerificationResult {
        num_accepted: 3,
        accepted_tokens: vec![10, 20, 30],
        bonus_token: Some(40),
        first_rejection_pos: -1,
    };

    assert_eq!(result.num_accepted, 3);
    assert_eq!(result.accepted_tokens.len(), 3);
    assert_eq!(result.bonus_token, Some(40));
}

#[test]
fn test_apply_temperature_and_softmax() {
    let draft = MockModel::new(4, "draft");
    let target = MockModel::new(4, "target");
    let config = SpeculativeConfig::default();
    let executor = SpeculativeExecutor::<CpuRuntime, _, _>::new(draft, target, config);

    let logits = vec![1.0, 2.0, 3.0, 4.0];
    let log_probs = executor.apply_temperature_and_softmax(&logits, 1.0);

    let sum: f32 = log_probs.iter().map(|&lp| lp.exp()).sum();
    assert!((sum - 1.0).abs() < 1e-5);

    assert!(log_probs[3] > log_probs[2]);
    assert!(log_probs[2] > log_probs[1]);
    assert!(log_probs[1] > log_probs[0]);
}

#[test]
fn test_adaptive_depth() {
    let draft = MockModel::new(100, "draft");
    let target = MockModel::new(100, "target");
    let config = SpeculativeConfig {
        num_speculative_tokens: 4,
        adaptive_depth: true,
        min_acceptance_rate: 0.3,
        ..Default::default()
    };

    let mut executor = SpeculativeExecutor::<CpuRuntime, _, _>::new(draft, target, config);

    executor.stats.accepted_tokens = 10;
    executor.stats.rejected_tokens = 90;
    executor.adjust_depth();
    assert!(executor.current_depth() < 4);

    executor.stats.accepted_tokens = 90;
    executor.stats.rejected_tokens = 10;
    executor.current_depth = 2;
    executor.adjust_depth();
    assert!(executor.current_depth() > 2);
}