Int8DynActInt4WeightQATQuantizer

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

Quantizer for performing QAT on a model, where linear layers have int8 dynamic per token fake quantized activations and int4 fake quantized grouped per channel weights.