float16_dynamic_qconfig

torch.ao.quantization.qconfig.float16_dynamic_qconfig

Dynamic qconfig with weights quantized to torch.float16.