Skip to content

training_compute_cost

mlte/qa_category/costs/training_compute_cost.py

TrainingComputeCost QACategory definition.

TrainingComputeCost

Bases: QACategory

The TrainingComputeCost QACategory reflects the costs of model training associated with compute resources.

Source code in mlte/qa_category/costs/training_compute_cost.py 
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
class TrainingComputeCost(QACategory):
    """
    The TrainingComputeCost QACategory reflects the costs
    of model training associated with compute resources.
    """

    def __init__(self, rationale: str):
        """
        Initialize a TrainingComputeCost instance.
        """
        super().__init__(
            instance=self,
            description="""
                The TrainingComputeCost QACategory assesses the
                computational requirements of model training.
                This might be measured in terms of CPU utilization
                for training processes that run locally, or the cost
                of compute resources required for training processes
                that run on on-demand cloud infrastructure.
                """,
            rationale=rationale,
        )

__init__(rationale)

Initialize a TrainingComputeCost instance.

Source code in mlte/qa_category/costs/training_compute_cost.py 
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
def __init__(self, rationale: str):
    """
    Initialize a TrainingComputeCost instance.
    """
    super().__init__(
        instance=self,
        description="""
            The TrainingComputeCost QACategory assesses the
            computational requirements of model training.
            This might be measured in terms of CPU utilization
            for training processes that run locally, or the cost
            of compute resources required for training processes
            that run on on-demand cloud infrastructure.
            """,
        rationale=rationale,
    )