Skip to content

Optimizer

Optimizer #

Base class. Implements all low-level functionality.

Parameters:

  • name (str, default: None ) –

    Name of the subdirectory inside path where model will be saved. The final model directory will be os.path.join(path, name) If None, defaults to the model's class name: self.class.name

  • path (str, default: None ) –

    Directory location to store all outputs. If None, a new unique time-stamped directory is chosen.

Source code in src/qtt/optimizers/optimizer.py 
 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
141
142
143
144
145
146
147
148
class Optimizer:
    """Base class. Implements all low-level functionality.

    Args:
        name (str):
            Name of the subdirectory inside path where model will be saved.
            The final model directory will be os.path.join(path, name)
            If None, defaults to the model's class name: self.__class__.__name__
        path (str):
            Directory location to store all outputs.
            If None, a new unique time-stamped directory is chosen.
    """

    model_file_name = "model.pkl"

    def __init__(
        self,
        name: str | None = None,
        path: str | None = None,
    ):
        if name is None:
            self.name = self.__class__.__name__
            logger.info(
                f"No name was specified for model, defaulting to class name: {self.name}",
            )
        else:
            self.name = name

        if path is None:
            self.path = setup_outputdir(path=self.name.lower())
            logger.info(
                f"No path was specified for predictor, defaulting to: {self.path}",
            )
        else:
            self.path = setup_outputdir(path)

        self._is_initialized = False

    def ante(self):
        """This method is intended for the use with a tuner.
        It allows to perform some pre-processing steps before each ask."""
        pass

    def post(self):
        """This method is intended for the use with a tuner.
        It allows to perform some post-processing steps after each tell."""
        pass

    def ask(self) -> dict:
        """Ask the optimizer for a trial to evaluate.

        Returns:
            A config to sample.
        """
        raise NotImplementedError

    def tell(self, report: dict | list[dict]):
        """Tell the optimizer the result for an asked trial.

        Args:
            report (dict): The result for a trial
        """
        raise NotImplementedError

    @classmethod
    def load(cls, path: str, reset_paths: bool = True, verbose: bool = True):
        """
        Loads the model from disk to memory.

        Args:
            path (str):
                Path to the saved model, minus the file name.
                This should generally be a directory path ending with a '/' character (or appropriate path separator value depending on OS).
                The model file is typically located in os.path.join(path, cls.model_file_name).
            reset_paths (bool):
                Whether to reset the self.path value of the loaded model to be equal to path.
                It is highly recommended to keep this value as True unless accessing the original self.path value is important.
                If False, the actual valid path and self.path may differ, leading to strange behaviour and potential exceptions if the model needs to load any other files at a later time.
            verbose (bool):
                Whether to log the location of the loaded file.

        Returns:
            model (Optimizer): The loaded model object.
        """
        file_path = os.path.join(path, cls.model_file_name)
        with open(file_path, "rb") as f:
            model = pickle.load(f)
        if reset_paths:
            model.path = path
        if verbose:
            logger.info(f"Model loaded from: {file_path}")
        return model

    def save(self, path: str | None = None, verbose: bool = True) -> str:
        """
        Saves the model to disk.

        Args:
            path (str): Path to the saved model, minus the file name. This should generally
                be a directory path ending with a '/' character (or appropriate path separator
                value depending on OS). If None, self.path is used. The final model file is
                typically saved to os.path.join(path, self.model_file_name).
            verbose (bool): Whether to log the location of the saved file.

        Returns:
            str: Path to the saved model, minus the file name. Use this value to load the
                model from disk via cls.load(path), where cls is the class of the model
                object (e.g., model = Model.load(path)).
        """
        if path is None:
            path = self.path
        os.makedirs(path, exist_ok=True)
        file_path = os.path.join(path, self.model_file_name)
        # tmp = {}
        # for key, obj in vars(self).items():
        #     if hasattr(obj, "save"):
        #         obj_path = os.path.join(path, key)
        #         obj.save(obj_path)
        #         tmp[key] = obj
        #         setattr(self, key, None)
        with open(file_path, "wb") as f:
            pickle.dump(self, f, protocol=pickle.HIGHEST_PROTOCOL)
        # for key, obj in tmp.items():
        #     setattr(self, key, obj)
        if verbose:
            logger.info(f"Model saved to: {file_path}")
        return path

    def reset_path(self, path: str | None = None):
        """Reset the path of the model.

        Args:
            path (str):
                Directory location to store all outputs.
                If None, a new unique time-stamped directory is chosen.
        """
        if path is None:
            path = setup_outputdir(path=self.name.lower(), path_suffix=self.name)
        self.path = path

ante() #

This method is intended for the use with a tuner. It allows to perform some pre-processing steps before each ask.

Source code in src/qtt/optimizers/optimizer.py 
48
49
50
51
def ante(self):
    """This method is intended for the use with a tuner.
    It allows to perform some pre-processing steps before each ask."""
    pass

ask() #

Ask the optimizer for a trial to evaluate.

Returns:

  • dict –

    A config to sample.

Source code in src/qtt/optimizers/optimizer.py 
58
59
60
61
62
63
64
def ask(self) -> dict:
    """Ask the optimizer for a trial to evaluate.

    Returns:
        A config to sample.
    """
    raise NotImplementedError

load(path, reset_paths=True, verbose=True) classmethod #

Loads the model from disk to memory.

Parameters:

  • path (str) –

    Path to the saved model, minus the file name. This should generally be a directory path ending with a '/' character (or appropriate path separator value depending on OS). The model file is typically located in os.path.join(path, cls.model_file_name).

  • reset_paths (bool, default: True ) –

    Whether to reset the self.path value of the loaded model to be equal to path. It is highly recommended to keep this value as True unless accessing the original self.path value is important. If False, the actual valid path and self.path may differ, leading to strange behaviour and potential exceptions if the model needs to load any other files at a later time.

  • verbose (bool, default: True ) –

    Whether to log the location of the loaded file.

Returns:

  • model ( Optimizer ) –

    The loaded model object.

Source code in src/qtt/optimizers/optimizer.py 
 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
@classmethod
def load(cls, path: str, reset_paths: bool = True, verbose: bool = True):
    """
    Loads the model from disk to memory.

    Args:
        path (str):
            Path to the saved model, minus the file name.
            This should generally be a directory path ending with a '/' character (or appropriate path separator value depending on OS).
            The model file is typically located in os.path.join(path, cls.model_file_name).
        reset_paths (bool):
            Whether to reset the self.path value of the loaded model to be equal to path.
            It is highly recommended to keep this value as True unless accessing the original self.path value is important.
            If False, the actual valid path and self.path may differ, leading to strange behaviour and potential exceptions if the model needs to load any other files at a later time.
        verbose (bool):
            Whether to log the location of the loaded file.

    Returns:
        model (Optimizer): The loaded model object.
    """
    file_path = os.path.join(path, cls.model_file_name)
    with open(file_path, "rb") as f:
        model = pickle.load(f)
    if reset_paths:
        model.path = path
    if verbose:
        logger.info(f"Model loaded from: {file_path}")
    return model

post() #

This method is intended for the use with a tuner. It allows to perform some post-processing steps after each tell.

Source code in src/qtt/optimizers/optimizer.py 
53
54
55
56
def post(self):
    """This method is intended for the use with a tuner.
    It allows to perform some post-processing steps after each tell."""
    pass

reset_path(path=None) #

Reset the path of the model.

Parameters:

  • path (str, default: None ) –

    Directory location to store all outputs. If None, a new unique time-stamped directory is chosen.

Source code in src/qtt/optimizers/optimizer.py 
138
139
140
141
142
143
144
145
146
147
148
def reset_path(self, path: str | None = None):
    """Reset the path of the model.

    Args:
        path (str):
            Directory location to store all outputs.
            If None, a new unique time-stamped directory is chosen.
    """
    if path is None:
        path = setup_outputdir(path=self.name.lower(), path_suffix=self.name)
    self.path = path

save(path=None, verbose=True) #

Saves the model to disk.

Parameters:

  • path (str, default: None ) –

    Path to the saved model, minus the file name. This should generally be a directory path ending with a '/' character (or appropriate path separator value depending on OS). If None, self.path is used. The final model file is typically saved to os.path.join(path, self.model_file_name).

  • verbose (bool, default: True ) –

    Whether to log the location of the saved file.

Returns:

  • str ( str ) –

    Path to the saved model, minus the file name. Use this value to load the model from disk via cls.load(path), where cls is the class of the model object (e.g., model = Model.load(path)).

Source code in src/qtt/optimizers/optimizer.py 
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
def save(self, path: str | None = None, verbose: bool = True) -> str:
    """
    Saves the model to disk.

    Args:
        path (str): Path to the saved model, minus the file name. This should generally
            be a directory path ending with a '/' character (or appropriate path separator
            value depending on OS). If None, self.path is used. The final model file is
            typically saved to os.path.join(path, self.model_file_name).
        verbose (bool): Whether to log the location of the saved file.

    Returns:
        str: Path to the saved model, minus the file name. Use this value to load the
            model from disk via cls.load(path), where cls is the class of the model
            object (e.g., model = Model.load(path)).
    """
    if path is None:
        path = self.path
    os.makedirs(path, exist_ok=True)
    file_path = os.path.join(path, self.model_file_name)
    # tmp = {}
    # for key, obj in vars(self).items():
    #     if hasattr(obj, "save"):
    #         obj_path = os.path.join(path, key)
    #         obj.save(obj_path)
    #         tmp[key] = obj
    #         setattr(self, key, None)
    with open(file_path, "wb") as f:
        pickle.dump(self, f, protocol=pickle.HIGHEST_PROTOCOL)
    # for key, obj in tmp.items():
    #     setattr(self, key, obj)
    if verbose:
        logger.info(f"Model saved to: {file_path}")
    return path

tell(report) #

Tell the optimizer the result for an asked trial.

Parameters:

  • report (dict) –

    The result for a trial

Source code in src/qtt/optimizers/optimizer.py 
66
67
68
69
70
71
72
def tell(self, report: dict | list[dict]):
    """Tell the optimizer the result for an asked trial.

    Args:
        report (dict): The result for a trial
    """
    raise NotImplementedError