tfm.optimization.LrConfig

Configuration for lr schedule.

Inherits From: OneOfConfig, Config, ParamsDict

tfm.optimization.LrConfig(
    default_params: dataclasses.InitVar[Optional[Mapping[str, Any]]] = None,
    restrictions: dataclasses.InitVar[Optional[List[str]]] = None,
    type: Optional[str] = None,
    constant: tfm.optimization.ConstantLrConfig = lr_cfg.ConstantLrConfig(),
    stepwise: tfm.optimization.StepwiseLrConfig = lr_cfg.StepwiseLrConfig(),
    exponential: tfm.optimization.ExponentialLrConfig = lr_cfg.ExponentialLrConfig(),
    polynomial: tfm.optimization.PolynomialLrConfig = lr_cfg.PolynomialLrConfig(),
    cosine: tfm.optimization.CosineLrConfig = lr_cfg.CosineLrConfig(),
    power: tfm.optimization.DirectPowerLrConfig = lr_cfg.DirectPowerLrConfig(),
    power_linear: tfm.optimization.PowerAndLinearDecayLrConfig = lr_cfg.PowerAndLinearDecayLrConfig(),
    power_with_offset: tfm.optimization.PowerDecayWithOffsetLrConfig = lr_cfg.PowerDecayWithOffsetLrConfig(),
    step_cosine_with_offset: tfm.optimization.StepCosineLrConfig = lr_cfg.StepCosineLrConfig()
)

Attributes
type	'str', type of lr schedule to be used, one of the fields below.
constant	constant learning rate config.
stepwise	stepwise learning rate config.
exponential	exponential learning rate config.
polynomial	polynomial learning rate config.
cosine	cosine learning rate config.
power	step^power learning rate config.
power_linear	learning rate config of step^power followed by step^power*linear.
power_with_offset	power decay with a step offset.
step_cosine_with_offset	Step cosine with a step offset.
BUILDER
default_params	Dataclass field
restrictions	Dataclass field

Methods

as_dict

View source

as_dict()

Returns a dict representation of OneOfConfig.

For the nested base_config.Config, a nested dict will be returned.

from_args

View source

@classmethod
from_args(
    *args, **kwargs
)

Builds a config from the given list of arguments.

from_json

View source

@classmethod
from_json(
    file_path: str
)

Wrapper for from_yaml.

from_yaml

View source

@classmethod
from_yaml(
    file_path: str
)

get

View source

get()

Returns selected config based on the value of type.

If type is not set (None), None is returned.

lock

View source

lock()

Makes the ParamsDict immutable.

override

View source

override(
    override_params, is_strict=True
)

Override the ParamsDict with a set of given params.

Args
override_params	a dict or a ParamsDict specifying the parameters to be overridden.
is_strict	a boolean specifying whether override is strict or not. If True, keys in override_params must be present in the ParamsDict. If False, keys in override_params can be different from what is currently defined in the ParamsDict. In this case, the ParamsDict will be extended to include the new keys.

replace

View source

replace(
    **kwargs
)

Overrides/returns a unlocked copy with the current config unchanged.

validate

View source

validate()

Validate the parameters consistency based on the restrictions.

This method validates the internal consistency using the pre-defined list of restrictions. A restriction is defined as a string which specfiies a binary operation. The supported binary operations are {'==', '!=', '<', '<=', '>', '>='}. Note that the meaning of these operators are consistent with the underlying Python immplementation. Users should make sure the define restrictions on their type make sense.

For example, for a ParamsDict like the following

a:
  a1: 1
  a2: 2
b:
  bb:
    bb1: 10
    bb2: 20
  ccc:
    a1: 1
    a3: 3

one can define two restrictions like this ['a.a1 == b.ccc.a1', 'a.a2 <= b.bb.bb2']

What it enforces are
a.a1 = 1 == b.ccc.a1 = 1 a.a2 = 2 <= b.bb.bb2 = 20

Raises
KeyError	if any of the following happens (1) any of parameters in any of restrictions is not defined in ParamsDict, (2) any inconsistency violating the restriction is found.
ValueError	if the restriction defined in the string is not supported.

__contains__

View source

__contains__(
    key
)

Implements the membership test operator.

__eq__

__eq__(
    other
)

Class Variables
IMMUTABLE_TYPES	(<class 'str'>, <class 'int'>, <class 'float'>, <class 'bool'>, <class 'NoneType'>)
RESERVED_ATTR	['_locked', '_restrictions']
SEQUENCE_TYPES	(<class 'list'>, <class 'tuple'>)
constant	Instance of tfm.optimization.ConstantLrConfig
cosine	Instance of tfm.optimization.CosineLrConfig
default_params	None
exponential	Instance of tfm.optimization.ExponentialLrConfig
polynomial	Instance of tfm.optimization.PolynomialLrConfig
power	Instance of tfm.optimization.DirectPowerLrConfig
power_linear	Instance of tfm.optimization.PowerAndLinearDecayLrConfig
power_with_offset	Instance of tfm.optimization.PowerDecayWithOffsetLrConfig
restrictions	None
step_cosine_with_offset	Instance of tfm.optimization.StepCosineLrConfig
stepwise	Instance of tfm.optimization.StepwiseLrConfig
type	None