dataquality.dq_start package#

Module contents#

class BaseInsights(model, *args, **kwargs)#

Bases: ABC

Base class for dq start integrations.

Initialize the base class. :type model: Any :param model: The model to be tracked :type args: Any :param args: Positional arguments to be passed to the watch function :type kwargs: Any :param kwargs: Keyword arguments to be passed to the watch function

framework: ModelFramework#
watch: Optional[Callable]#
unwatch: Optional[Callable]#
call_finish: bool = True#
enter()#

Call the watch function (called in __enter__).

Return type:

None

exit()#

Call the unwatch function (called in __exit__).

Return type:

None

set_project_run(project='', run='', task=TaskType.text_classification)#

Set the project and run names. To the class. If project and run are not provided, generate them. :type project: str :param project: The project name :type run: str :param run: The run name :type task: TaskType :param task: The task type

Return type:

None

init_project(task, project='', run='')#

Initialize the project and calls dq.init(). :type task: TaskType :param task: The task type :type project: str :param project: The project name :type run: str :param run: The run name

Return type:

None

setup_training(labels, train_data, test_data=None, val_data=None)#

Log dataset and labels to the run. :type labels: Optional[List[str]] :param labels: The labels :type train_data: Any :param train_data: The training dataset :type test_data: Optional[Any] :param test_data: The test dataset :type val_data: Optional[Any] :param val_data: The validation dataset

Return type:

None

validate(task_type, labels)#

Validate the task type and labels. :type task_type: TaskType :param task_type: The task type :type labels: Optional[List[str]] :param labels: The labels

Return type:

None

class TorchInsights(model)#

Bases: BaseInsights

Initialize the base class. :type model: Any :param model: The model to be tracked :param args: Positional arguments to be passed to the watch function :param kwargs: Keyword arguments to be passed to the watch function

framework: ModelFramework = 'torch'#
class TFInsights(model)#

Bases: BaseInsights

Initialize the base class. :type model: Any :param model: The model to be tracked :param args: Positional arguments to be passed to the watch function :param kwargs: Keyword arguments to be passed to the watch function

framework: ModelFramework = 'keras'#
class TrainerInsights(model)#

Bases: BaseInsights

Initialize the base class. :type model: Any :param model: The model to be tracked :param args: Positional arguments to be passed to the watch function :param kwargs: Keyword arguments to be passed to the watch function

framework: ModelFramework = 'hf'#
class AutoInsights(model)#

Bases: BaseInsights

Initialize the base class. :type model: Any :param model: The model to be tracked :param args: Positional arguments to be passed to the watch function :param kwargs: Keyword arguments to be passed to the watch function

framework: ModelFramework = 'auto'#
call_finish: bool = False#
auto_kwargs: Dict[str, Any]#
setup_training(labels, train_data, test_data=None, val_data=None)#

Setup auto by creating the parameters for the auto function. :type labels: Optional[List[str]] :param labels: Labels for the training :type train_data: Any :param train_data: Training dataset :type test_data: Optional[Any] :param test_data: Test dataset :type val_data: Optional[Any] :param val_data: Validation dataset

Return type:

None

init_project(task, project='', run='')#

Initialize the project and run but dq init is not called. :type task: TaskType :param task: The task type :type project: str :param project: The project name :type run: str :param run: The run name

Return type:

None

enter()#

Call auto function with the generated paramters.

Return type:

None

validate(task_type, labels=[])#

Validate the task type and labels. :type task_type: TaskType :param task_type: The task type :type labels: Optional[List[str]] :param labels: The labels

Return type:

None

detect_model(model, framework)#

Detect the model type in a lazy way and return the appropriate class. :type model: Any :param model: The model to inspect, if a string, it will be assumed to be auto :rtype: Type[BaseInsights]

Parameters:

framework (Optional[ModelFramework]) – The framework to use, if provided it will be used instead of the model

class DataQuality(model=None, task=TaskType.text_classification, labels=None, train_data=None, test_data=None, val_data=None, project='', run='', framework=None, *args, **kwargs)#

Bases: object

Parameters:

  • model (Optional[Any]) – The model to inspect, if a string, it will be assumed to be auto

  • task (TaskType) – Task type for example “text_classification”

  • project (str) – Project name

  • run (str) – Run name

  • train_data (Optional[Any]) – Training data

  • test_data (Optional[Any]) – Optional test data

  • val_data (Optional[Any]) – Optional: validation data

  • labels (Optional[List[str]]) – The labels for the run

  • framework (Optional[ModelFramework]) – The framework to use, if provided it will be used instead of inferring it from the model. For example, if you have a torch model, you can pass framework=”torch”. If you have a torch model, you can pass framework=”torch”

  • args (Any) – Additional arguments

  • kwargs (Any) – Additional keyword arguments

from dataquality import DataQuality

with DataQuality(model, "text_classification",
                 labels = ["neg", "pos"],
                 train_data = train_data) as dq:
    model.fit(train_data)

If you want to train without a model, you can use the auto framework:

from dataquality import DataQuality

with DataQuality(labels = ["neg", "pos"],
                 train_data = train_data) as dq:
    dq.finish()
get_metrics(split=Split.train)#
Return type:

Dict[str, Any]