It's now 2026 for me so I guess I'll do a New Years update post since everyone else is since I've had entirely too much cider (Meaning unfiltered apple juice with artificial carbonation. Normally don't have several hundred calories of sugar this late.).
For 2026 I've already got 3 new loras trained scheduled and an almost complete new dataset I just need to adjust tags on to throw in the trainer. Also got an existing LoRA I made a natural language dataset version of for for practice, and a requested Pony conversion. I am not going to repeat the insanity of last year's daily LoRAs for 8 months, then another set of daily LoRAs for 25 days at Christmas, but I hope LoRAs that come primarily when interest, will, and dataset existence all align will continue to interest everyone! Hopefully shouldn't be that rare, since I do look at stuff and wonder "I wonder if there's enough of that for a LoRA" a lot.
The stuff I mentioned above will put me at 190 distinct LoRAs, or 192 counting the new natural language datasets for existing works as their own LoRA (my poll on the subject got inconclusive results). I'm already working on my 200th LoRA, but it will have more pictures in its dataset than any non-style I've done (that personal record is held by my Cindy Vortex LoRA at just under 200): I haven't captured all the data I'll need for LoRA #200 yet, but just my finished and to clean folder already hit over that (though several of those are going to be discarded as redundant or uneeded).
Pic unrelated. It's just what I had in generator and looked nice enough.