The code flow is slightly different from the paper？

[pengpenglove]

In paper，The weight after each quantization will not be changed in the subsequent retraining and quantization process of expanding the proportion
In code, When the proportion of the last two quantization is 50%, the quantized weight will be put into the second group again in the next proportional partition, which will be propagated back in the retraining.
My understanding is correct?