Skip to content

Module: vector

DataType

Bases: IntEnum

Supported data types

Source code in src/infernet_ml/utils/codec/vector.py 
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
class DataType(IntEnum):
    """Supported data types"""

    float = 0
    double = 1
    cfloat = 2
    cdouble = 3
    half = 4
    bfloat16 = 5
    uint8 = 6
    int8 = 7
    short = 8
    int = 9
    long = 10
    bool = 11

encode_vector(dtype, shape, values)

Shape of the tensor will be flattened.

Source code in src/infernet_ml/utils/codec/vector.py 
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
def encode_vector(
    dtype: DataType, shape: Tuple[int, ...], values: Union[Tensor, List[Any]]
) -> bytes:
    """
    Shape of the tensor will be flattened.
    """
    if isinstance(values, Tensor):
        _values = values.flatten().tolist()
    else:
        _values = np.array(values).flatten().tolist()

    map_fn_lookup = {
        DataType.float: float_to_u32,
        DataType.double: double_to_u64,
        DataType.cfloat: cfloat_to_u64,
        DataType.cdouble: cdouble_to_u128,
        DataType.half: half_to_u16,
        DataType.bfloat16: bfloat16_to_u16,
        DataType.uint8: lambda x: x,
        DataType.int8: lambda x: x,
        DataType.short: lambda x: x,
        DataType.int: lambda x: x,
        DataType.long: lambda x: x,
        DataType.bool: lambda x: x,
    }
    __values = _map(map_fn_lookup[dtype], _values)

    _type = SOLIDITY_TYPE_LOOKUP[dtype]

    return encode(
        ["uint8", "uint16[]", f"{_type}[]"],
        [dtype, shape, __values],
    )