Int8DynActInt4WeightQATQuantizer

class torchao.quantization.qat.Int8DynActInt4WeightQATQuantizer(groupsize: int = 256, padding_allowed: bool = False, precision: dtype = torch.float32, scales_precision: dtype = torch.float32)

执行 QAT 的量化器，其中线性层具有 int8 动态每 token 假量化激活和 int4 假量化每通道分组权重。