issue about speaker embedding

[zhenye234]

Hello, thanks for your great work.

so-vits-svc/models.py

Line 437 in 4078cb7

z_ptemp, m_p, logs_p, _ = self.enc_p(x, x_mask, f0=f0_to_coarse(f0))

I am confused about why not add speaker embedding for z_p prediction. Because the z has used the speaker embedding and z_p should also include this information. May I know why ?