Do you think editing the name of this model's safetensor file before placing the model in the LoRA folder would negate this issue? Thanks for the heads up. I'll try to comment back with whatever answer I get from trying that approach.

@DevilDogAV That would work, I'd say (SD.Next has adjusted the parsing, but there's no way to tell if it wouldn't break past metadata).

@einar_rainhart I asked AI and it confirmed that it could be a problem. AI said my solution should totally prevent the issue. I downloaded the model and just erased the unnecessary characters in the title of the safetensors file (most importantly "steps"). I've been able to use the LoRA on Automatic1111 without issue. The LoRA works quite well, I'm sure many experienced SD artists are aware of how overexposed images can get at times. This resource seems like it can be a real asset. I recommend the author just re-upload the model as a v2 so that they can keep this listing and all their thumbs-ups, stats, ect, and then just make v1 unavailable for download. It seems like a great addition to one's kit.

@DevilDogAV As I said, at least for SD.Next the developer changed the parsing code and it's no longer an issue, but for A1111 it could be still, indeed.

@Mraclet Who would know that would be an issue. hehehe Such a random coincidence for something so innocuous to be a potential issue. Shout out to @einar_rainhart for making us aware and props to you for addressing it. Please confirm that my suggestion will preserve your listing page here, it would be a shame to lose all the stats you've earned from this model. If in doubt, contact support, I'm sure they can help you resolve the issue in a way that everybody wins. Extremely useful model! Thanks for creating it.

@Mraclet it really is, it's giving my images a depth and style I haven't been able to capture before, without intruding into the actual generation itself, it's incredible. I just started using it yesterday and I think it will be part of all my workflows from now on.