MarlinQQQTensor

class torchao.dtypes.MarlinQQQTensor(tensor_impl: AQTTensorImpl, block_size: Tuple[int, ...], shape: Size, quant_min: Optional[Union[int, float]] = None, quant_max: Optional[Union[int, float]] = None, zero_point_domain: ZeroPointDomain = ZeroPointDomain.INT, dtype=None, strides=None)

MarlinQQQ 量化張量子類，繼承自 AffineQuantizedTensor 類。

要了解 _choose_qparams_and_quantize_affine_qqq、Marlin QQQ 量化的量化和反量化過程中發生的情況，請檢視 https://github.com/pytorch/ao/blob/main/torchao/quantization/quant_primitives.py 並檢查兩個量化原始操作：_choose_qparams_and_quantize_affine_qqq 和 _dequantize_affine_qqq

dequantize() → Tensor

給定一個量化張量，對其進行去量化並返回去量化的浮點張量。

classmethod from_hp_to_intx(input_float: Tensor, block_size: Tuple[int, ...], quant_min: Optional[int] = None, quant_max: Optional[int] = None, zero_point_domain: ZeroPointDomain = ZeroPointDomain.INT, _layout: Optional[Layout] = None)

將浮點張量轉換為 Marlin QQQ 量化張量。