FixedQParamsFakeQuantize

class torch.quantization.fake_quantize.FixedQParamsFakeQuantize(scale, zero_point, dtype=torch.quint8, qscheme=torch.per_tensor_affine, quant_min=0, quant_max=255)

Simulate quantize and dequantize with fixed quantization parameters in training time. Only per tensor quantization is supported.

Parameters

• scale (float) – fixed scale for the fake quantize module

• zero_point (int) – fixed zero point for the fake quantize module

• dtype –

• qscheme –

• quant_min –

• quant_max –