mlos_bench.tunables.tunable_types

Helper types for Tunable.

The main class of interest to most users in this module is TunableDict, which provides the typed conversions from a JSON config to a config used for creating a Tunable.

The other types are mostly used for type checking and documentation purposes.

Attributes

`DistributionName`	The `distribution` type names for a `Tunable` value.
`TUNABLE_DTYPE`	Maps `Tunable` types to their corresponding Python data types by name.
`TunableValue`	A `TypeAlias` for a `Tunable` parameter value.
`TunableValueType`	A `TypeAlias` for `Tunable` value
`TunableValueTypeName`	The accepted string names of a `Tunable` value `type`.
`TunableValueTypeTuple`	Tunable value `type` tuple.
`TunableValuesDict`	Tunable values dictionary type.

Classes

`DistributionDict`	A `TypedDict` for a `Tunable` parameter's required
`DistributionDictOpt`	A `TypedDict` for a `Tunable` parameter's optional
`TunableDict`	A `TypedDict` for a `Tunable` parameter's required config
`TunableDictOpt`	A `TypedDict` for a `Tunable` parameter's optional config

Functions

tunable_dict_from_dict(→ TunableDict)

Creates a TunableDict from a regular dict.

Module Contents

class mlos_bench.tunables.tunable_types.DistributionDict(*args, **kwargs)[source]

Bases: DistributionDictOpt

A TypedDict for a Tunable parameter’s required distribution config parameters.

Mostly used for type checking. These are the types expected to be received from the json config.

See also

Tunable.distribution :: Examples of Tunables with distributions.
Tunable.distribution_params :: Examples of distribution parameters.

type: DistributionName[source]

The name of the distribution.

See also

Tunable.distribution :: Examples of distribution names.

class mlos_bench.tunables.tunable_types.DistributionDictOpt(*args, **kwargs)[source]

Bases: TypedDict

A TypedDict for a Tunable parameter’s optional distribution_params config.

Mostly used for type checking. These are the types expected to be received from the json config.

Notes

DistributionDict contains the required fields for the Tunable.distribution parameter.

See also

Tunable.distribution_params :: Examples of distribution parameters.

params: dict[str, float] | None[source]

The parameters for the distribution.

See also

Tunable.distribution_params :: Examples of distribution parameters.

class mlos_bench.tunables.tunable_types.TunableDict(*args, **kwargs)[source]

Bases: TunableDictOpt

A TypedDict for a Tunable parameter’s required config parameters.

Mostly used for type checking. These are the types expected to be received from the json config.

Examples

>>> # Example values of the TunableDict
>>> TunableDict({'type': 'int', 'default': 0, 'range': [0, 10]})
{'type': 'int', 'default': 0, 'range': [0, 10]}

>>> # Example values of the TunableDict with optional fields
>>> TunableDict({'type': 'categorical', 'default': 'a', 'values': ['a', 'b']})
{'type': 'categorical', 'default': 'a', 'values': ['a', 'b']}

default: TunableValue[source]: The default value of the Tunable parameter.

See also

Tunable.default

type: TunableValueTypeName[source]: The name of the type of the Tunable parameter.

See also

Tunable.type : Examples of type names.

class mlos_bench.tunables.tunable_types.TunableDictOpt(*args, **kwargs)[source]

Bases: TypedDict

A TypedDict for a Tunable parameter’s optional config parameters.

Mostly used for type checking. These are the types expected to be received from the json config.

Notes

TunableDict contains the required fields for the Tunable parameter.

description: str | None[source]: Description of the Tunable parameter.

See also

Tunable.description

distribution: DistributionDict | None[source]

Optional sampling distribution configuration for an “int” or “float” type Tunable parameter.

See also

Tunable.distribution :: Examples of distributions.
Tunable.distribution_params :: Examples of distribution parameters.

log: bool | None[source]: Whether to use log sampling for an “int” or “float” type Tunable parameter.

See also

Tunable.is_log

meta: dict[str, Any][source]: Free form dict to store additional metadata for the Tunable parameter (e.g., unit suffix, etc.)

See also

Tunable.meta : Examples of Tunables with metadata.

quantization_bins: int | None[source]

The number of quantization bins for an “int” or “float” type Tunable parameter.

See also

Tunable.quantization_bins :: Examples of quantized Tunables.

range: collections.abc.Sequence[int] | collections.abc.Sequence[float] | None[source]

The range of values for an “int” or “float” type Tunable parameter.

Must be a sequence of two values: [min, max].

A range is required for “int” and “float” type Tunables.

See also

Tunable.range : Examples of ranges. Tunable.values

range_weight: float | None[source]

Optional sampling weight for the main ranges of an “int” or “float” type Tunable parameter.

See also

Tunable.range_weight :: Examples of weighted sampling Tunables.

special: list[int] | list[float] | None[source]

List of special values for an “int” or “float” type Tunable parameter.

These are values that are considered special by the target system (e.g., null, 0, -1, auto, etc.) and should be sampled with higher weights.

See also

Tunable.special : Examples of special values.

special_weights: list[float] | None[source]: Optional sampling weights for the special values of an “int” or “float” type Tunable parameter.

See also

Tunable.weights : Examples of weighted sampling Tunables.

values: list[str | None] | None[source]

List of values (or categories) for a “categorical” type Tunable parameter.

A list of values is required for “categorical” type Tunables.

See also

Tunable.categories Tunable.values

values_weights: list[float] | None[source]: Optional sampling weights for the values of a “categorical” type Tunable parameter.

See also

Tunable.weights : Examples of weighted sampling Tunables.

mlos_bench.tunables.tunable_types.tunable_dict_from_dict(config: dict[str, Any]) → TunableDict[source]

Creates a TunableDict from a regular dict.

Notes

Mostly used for type checking while instantiating a Tunable from a json config.

Parameters:: config (dict[str, Any]) – A regular dict that represents a TunableDict.
Return type:: TunableDict

Examples

>>> # Example values of the TunableDict
>>> import json5 as json
>>> config = json.loads("{'type': 'int', 'default': 0, 'range': [0, 10]}")
>>> config
{'type': 'int', 'default': 0, 'range': [0, 10]}
>>> typed_dict = tunable_dict_from_dict(config)
>>> typed_dict
{'type': 'int', 'description': None, 'default': 0, 'values': None, 'range': [0, 10], 'quantization_bins': None, 'log': None, 'distribution': None, 'special': None, 'values_weights': None, 'special_weights': None, 'range_weight': None, 'meta': {}}

mlos_bench.tunables.tunable_types.DistributionName[source]

The distribution type names for a Tunable value.

See also

Tunable.distribution :: Example of accepted distribution names.

mlos_bench.tunables.tunable_types.TUNABLE_DTYPE: dict[TunableValueTypeName, TunableValueType][source]: Maps Tunable types to their corresponding Python data types by name.

See also

Tunable.dtype : Example of type mappings.

type mlos_bench.tunables.tunable_types.TunableValue = int | float | str | None[source]: A TypeAlias for a Tunable parameter value.

type mlos_bench.tunables.tunable_types.TunableValueType = type[int] | type[float] | type[str][source]: A TypeAlias for Tunable value data type.

See also

Tunable.dtype : Example of accepted types.

mlos_bench.tunables.tunable_types.TunableValueTypeName[source]: The accepted string names of a Tunable value type.

See also

Tunable.type : Example of accepted type names.

mlos_bench.tunables.tunable_types.TunableValueTypeTuple[source]

Tunable value type tuple.

Notes

For checking whether a param is a TunableValue with isinstance().

mlos_bench.tunables.tunable_types.TunableValuesDict[source]: Tunable values dictionary type.