Linear

class torch.ao.nn.quantized.Linear(in_features, out_features, bias_=True, dtype=torch.qint8)

A quantized linear module with quantized tensor as inputs and outputs. We adopt the same interface as torch.nn.Linear, please see https://pytorch.org/docs/stable/nn.html#torch.nn.Linear for documentation.

Similar to Linear, attributes will be randomly initialized at module creation time and will be overwritten later

Variables

• weight (Tensor) – the non-learnable quantized weights of the module of shape $(\text{out\_features}, \text{in\_features})$.

• bias (Tensor) – the non-learnable bias of the module of shape $(\text{out\_features})$. If bias is True, the values are initialized to zero.

• scale – scale parameter of output Quantized Tensor, type: double

• zero_point – zero_point parameter for output Quantized Tensor, type: long

Examples:

>>> m = nn.quantized.Linear(20, 30)
>>> input = torch.randn(128, 20)
>>> input = torch.quantize_per_tensor(input, 1.0, 0, torch.quint8)
>>> output = m(input)
>>> print(output.size())
torch.Size([128, 30])

classmethod from_float(mod, use_precomputed_fake_quant=False)

Create a quantized module from an observed float module

Parameters

• mod (Module) – a float module, either produced by torch.ao.quantization utilities or provided by the user

• use_precomputed_fake_quant (bool) – if True, the module will reuse min/max values from the precomputed fake quant module.

classmethod from_reference(ref_qlinear, output_scale, output_zero_point)

Create a (fbgemm/qnnpack) quantized module from a reference quantized module

Parameters

• ref_qlinear (Module) – a reference quantized linear module, either produced by torch.ao.quantization utilities or provided by the user

• output_scale (float) – scale for output Tensor

• output_zero_point (int) – zero point for output Tensor