quark.torch.quantization.nn.modules.quantize_linear

Module Contents

Classes

class quark.torch.quantization.nn.modules.quantize_linear.QuantLinear(in_features: int, out_features: int, device: torch.device, bias: bool, quant_config: quark.torch.quantization.config.config.QuantizationConfig, **kwargs: Any)

Quantized version of nn.Linear