llmcompressor.pipelines.independent

@CalibrationPipeline.register("independent")
class IndependentPipeline(CalibrationPipeline):
    @staticmethod
    def __call__(
        model: torch.nn.Module,
        dataloader: DataLoader,
        dataset_args: "DatasetArguments",
    ):
        """
        Data pipeline where each modifier is assigned its own calibration epoch and data
        pipeline

        :param model: model being calibrated
        :param dataloader: loads data for calibration
        :param dataset_args: dataset arguments relevant to pipelines
        """
        _logger = logger.patch(lambda r: r.update(function="IndependentPipeline"))

        session = active_session()
        modifiers = session.get_modifiers()
        with patch_attr(session.lifecycle, "modifiers", None):
            for index, modifier in enumerate(modifiers):
                mod_type = str(type(modifier).__name__)
                session.lifecycle.modifiers = [
                    StageModifiers(modifiers=[modifier], group=mod_type, index=index)
                ]

                pipeline = CalibrationPipeline.from_modifiers([modifier])
                pipeline_name = pipeline.__class__.__name__
                _logger.info(f"Inferred `{pipeline_name}` for `{mod_type}`")

                pipeline(model, dataloader, dataset_args)