Civit's search issue with pages of Loras that have different versions trained on different base models

Let’s say I trained style Lora of "NotvanGogh1337” using PonyXL as a base model. I upload the Lora and now people can find my Lora’s page if they search NotvanGogh1337 and check “Pony” as a “Base model” search filter.

Then, I decide to train the same Lora but for Illustrious. I do so, using Illustrious as base model and I upload NotvanGogh1337 Lora trained on Illustrious to the same page that I uploaded Pony version to. Now I have page of my NotvanGogh1337 style Lora and there is 2 different versions of it, trained on different base models.

Here is the issue: now, after I uploaded the Illustrious version and it is the latest one, people no longer can find my Lora’s page if they search NotvanGogh1337 and check “Pony” as a “Base model” search filter, there simply won’t be such searching result since the latest version of my lora was trained on Illustrious and so I set “Illustrious” on the model page as a Base model. Now they only can find my Lora’s page if they check Illustrious as a “Base model” search filter, even though the page of my Lora has Pony version as well.

Theoretical solution to that is to upload Loras trained on different base model to different pages, but that’s only theoretical, because this will overload the number of requests issued and generally is a crutch, not a solution to the problem. Why would we have such option as uploading different versions trained on different checkpoint to the same page if those are not properly indexed by Civit search?

I consider this as a serious issue since because of this the reach and recognition of the Loras decreases.

I apologize for bad English and a lot of repetition in the text above, I tried to clarify the issue as good as I could.

I hope you will process my request as soon as possible.