Method use_weight_only

tensorrt_llm/quantization/mode.py:356–362

(use_int4_weights=False, per_group=False)

    @staticmethod
    def use_weight_only(use_int4_weights=False, per_group=False):
        return QuantMode.from_description(quantize_weights=True,
                                          quantize_activations=False,
                                          per_token=False,
                                          per_channel=False,
                                          per_group=per_group,
                                          use_int4_weights=use_int4_weights)

    @staticmethod
    def from_quant_algo(