Training embedding with Hypernetwork loaded

[M-P-NaCl]

While training a new embedding I notice that I forgot to unload a Hypernetwork and was training with the Hypernetwork. Does this affect the training result? If it does, does that mean this embedding could only be used with that specific Hypernetwork?

[fredconex]

I was wondering the same, but so far I haven't noticed any differences so my guess is that it does not affect it, but would be great to have a word from @AUTOMATIC1111