Hi, if I quant the LSTM uing torch.ao.nn.quantized.LSTM and torch.ao.nn.quantizable.LSTM like pytorch/test/quantization/fx/test_quantize_fx.py at main · pytorch/pytorch · GitHub, how to get the scale and zp of LSTM hidden input h0 and c0? I want to quant the int8 model using long sequance data and inference it frame by frame, so I need to know the scale and zp of the hidden input of LSTM for every frame when inferencing.