aprender-registry 0.41.0

Model, Data and Recipe Registry with full lineage tracking
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

//! Lineage Tracking Example
//!
//! Demonstrates model lineage tracking:
//! - Fine-tuning relationships
//! - Quantization tracking
//! - Model merging
//!
//! Run with: cargo run --example lineage_tracking

use pacha::prelude::*;

fn main() {
    println!("=== Lineage Tracking Example ===\n");

    // Create a lineage graph
    let mut graph = LineageGraph::new();

    // Add base model
    println!("1. Creating base model node...");
    let base_id = ModelId::new();
    let base_idx = graph.add_node(pacha::lineage::LineageNode {
        model_id: base_id.clone(),
        model_name: "llama-7b".to_string(),
        model_version: "1.0.0".to_string(),
    });
    println!("   ✓ Base model: llama-7b:1.0.0");

    // Add fine-tuned model
    println!("\n2. Creating fine-tuned model node...");
    let finetuned_id = ModelId::new();
    let finetuned_idx = graph.add_node(pacha::lineage::LineageNode {
        model_id: finetuned_id.clone(),
        model_name: "llama-7b-fraud".to_string(),
        model_version: "1.0.0".to_string(),
    });

    // Add fine-tuning edge
    graph.add_edge(
        base_idx,
        finetuned_idx,
        ModelLineageEdge::FineTuned {
            parent: base_id.clone(),
            recipe: pacha::recipe::RecipeId::new(),
        },
    );
    println!("   ✓ Fine-tuned: llama-7b-fraud:1.0.0");
    println!("   ✓ Edge: llama-7b → llama-7b-fraud (fine-tuned)");

    // Add quantized model
    println!("\n3. Creating quantized model node...");
    let quantized_id = ModelId::new();
    let quantized_idx = graph.add_node(pacha::lineage::LineageNode {
        model_id: quantized_id.clone(),
        model_name: "llama-7b-fraud-int8".to_string(),
        model_version: "1.0.0".to_string(),
    });

    // Add quantization edge
    graph.add_edge(
        finetuned_idx,
        quantized_idx,
        ModelLineageEdge::Quantized {
            source: finetuned_id.clone(),
            quantization: QuantizationType::Int8,
        },
    );
    println!("   ✓ Quantized: llama-7b-fraud-int8:1.0.0");
    println!("   ✓ Edge: llama-7b-fraud → llama-7b-fraud-int8 (quantized INT8)");

    // Add another fine-tuned variant for merging
    println!("\n4. Creating another fine-tuned variant...");
    let variant_id = ModelId::new();
    let variant_idx = graph.add_node(pacha::lineage::LineageNode {
        model_id: variant_id.clone(),
        model_name: "llama-7b-code".to_string(),
        model_version: "1.0.0".to_string(),
    });

    graph.add_edge(
        base_idx,
        variant_idx,
        ModelLineageEdge::FineTuned {
            parent: base_id.clone(),
            recipe: pacha::recipe::RecipeId::new(),
        },
    );
    println!("   ✓ Fine-tuned: llama-7b-code:1.0.0");

    // Create merged model
    println!("\n5. Creating merged model...");
    let merged_id = ModelId::new();
    let merged_idx = graph.add_node(pacha::lineage::LineageNode {
        model_id: merged_id,
        model_name: "llama-7b-fraud-code".to_string(),
        model_version: "1.0.0".to_string(),
    });

    graph.add_edge(
        finetuned_idx,
        merged_idx,
        ModelLineageEdge::Merged {
            sources: vec![finetuned_id, variant_id],
            weights: vec![0.7, 0.3],
        },
    );
    println!("   ✓ Merged: llama-7b-fraud-code:1.0.0");
    println!("   ✓ Edge: llama-7b-fraud + llama-7b-code → merged (0.7/0.3)");

    // Print graph statistics
    println!("\n6. Lineage graph statistics:");
    println!("   Nodes: {}", graph.node_count());
    println!("   Edges: {}", graph.edge_count());

    // Show relationships
    println!("\n7. Model relationships:");
    println!("   Base model (idx={base_idx}):");
    println!("     Descendants: {:?}", graph.descendants(base_idx));

    println!("   Fine-tuned model (idx={finetuned_idx}):");
    println!("     Ancestors: {:?}", graph.ancestors(finetuned_idx));
    println!("     Descendants: {:?}", graph.descendants(finetuned_idx));

    println!("   Quantized model (idx={quantized_idx}):");
    println!("     Ancestors: {:?}", graph.ancestors(quantized_idx));

    // Demonstrate quantization types
    println!("\n8. Supported quantization types:");
    let quant_types = [
        QuantizationType::Int8,
        QuantizationType::Int4,
        QuantizationType::Fp16,
        QuantizationType::Bf16,
        QuantizationType::Dynamic,
    ];
    for qt in quant_types {
        println!("   - {qt}");
    }

    println!("\n✅ Lineage tracking example complete!");
}