onnx-ir 0.21.0

ONNX-IR is a pure Rust library for parsing ONNX models into an intermediate representation that can be used to generate code for various ML/DL frameworks
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

#!/usr/bin/env -S uv run
# /// script
# dependencies = [
#   "onnx>=1.15.0",
#   "numpy>=1.24.0",
# ]
# ///

"""
Generate ONNX model testing constant handling.

Tests:
- Initializers → Constant nodes (Static value source)
- Runtime inputs → Dynamic value source
- Constant lifting to Static arguments
- Unreferenced constant removal
"""

import onnx
from onnx import helper, TensorProto
import numpy as np


def create_constants_model():
    """Create model with various constant handling scenarios."""

    # Runtime input (Dynamic)
    runtime_input = helper.make_tensor_value_info("x", TensorProto.FLOAT, [1, 3, 4, 4])

    # Output
    output = helper.make_tensor_value_info("output", TensorProto.FLOAT, [1, 3, 4, 4])

    # Initializers (will become Constant nodes, then Static)
    # Used constant - for Add
    add_bias = helper.make_tensor(
        name="add_bias",
        data_type=TensorProto.FLOAT,
        dims=[1, 3, 1, 1],
        vals=np.array([1.0, 2.0, 3.0], dtype=np.float32).tobytes(),
        raw=True,
    )

    # Used constant - for Mul
    mul_scale = helper.make_tensor(
        name="mul_scale",
        data_type=TensorProto.FLOAT,
        dims=[1, 3, 1, 1],
        vals=np.array([0.5, 0.5, 0.5], dtype=np.float32).tobytes(),
        raw=True,
    )

    # Unused constant (should be removed in Phase 5)
    unused_const = helper.make_tensor(
        name="unused_constant",
        data_type=TensorProto.FLOAT,
        dims=[1, 3, 1, 1],
        vals=np.array([99.0, 99.0, 99.0], dtype=np.float32).tobytes(),
        raw=True,
    )

    # Create nodes
    nodes = [
        # Add with initializer (constant will be lifted to Static)
        helper.make_node("Add", ["x", "add_bias"], ["added"], name="add"),
        # Multiply with initializer
        helper.make_node("Mul", ["added", "mul_scale"], ["output"], name="mul"),
    ]

    # Create the graph
    graph = helper.make_graph(
        nodes,
        "constants_model",
        [runtime_input],
        [output],
        initializer=[
            add_bias,
            mul_scale,
            unused_const,
        ],  # unused_const should be removed
    )

    # Create the model
    model = helper.make_model(
        graph, producer_name="onnx-ir-test", opset_imports=[helper.make_opsetid("", 16)]
    )

    # Check the model
    onnx.checker.check_model(model)

    return model


def main():
    """Generate and save the ONNX model."""
    model = create_constants_model()

    # Save the model
    output_path = "../fixtures/constants.onnx"
    onnx.save(model, output_path)
    print(f"Model saved to {output_path}")

    # Print model info
    print(f"\nModel info:")
    print(f"  Opset version: {model.opset_import[0].version}")
    print(f"  Inputs: {[inp.name for inp in model.graph.input]}")
    print(f"  Outputs: {[out.name for out in model.graph.output]}")
    print(f"  Initializers: {[init.name for init in model.graph.initializer]}")
    print(f"  Nodes: {len(model.graph.node)}")
    for node in model.graph.node:
        print(
            f"    - {node.op_type} ({node.name}): {list(node.input)} → {list(node.output)}"
        )


if __name__ == "__main__":
    main()