This is AMAZING. Thank you very much for your wonderful work. I already know how to train styles on LoRA, but I don't know how to make LoRA like Enhancer that helps with this general realistically output. Can I get a little hint if possible? Also, what did you use for the training images? Thank you again. 👍👍

I'm loving this Lora! Already used it for multiple Deforum/ControlNet YT clips:

The Girl Crusader: https://youtu.be/jJsSM7_W3RA

Soulmates: https://youtu.be/xAp-wt3u9X4

Eppur si muove! https://youtu.be/cQNPBxe-EQk

This is certainly impressive. Thanks.

Works very well, thank you

Realy cool Lora!

Recomended; This Lora loves to add a fantasy landscape to the bottom of the images if you don't specify the environment. Great for character art, but if you need clean backgrounds you may want to turn this down a little or be very specific about the BG. I really like the details it adds, would be nice if it had more poses as it tends to put your subjects into the same pose and composition. Great with controlnets and ultimate upscaling.

Very good lora if you are looking to add detail to any type of image.

This LoRa might lead to multiple necks when drawing dragons, especially Chinese dragons.

I tried a couple of times with the same seed with values from 0 to 0.8 and it just makes my images worse...?!

I don't know what you put in this .safetensor (or model,) but roughly 90% of time when generating on site, it's equipping your LoRA or die of boredom. I find SDXL boring AF, the obsession for realism is ruining it - everything looks just the same, and stuck in time, like 1990s or earlier. Quite hard getting closer to at least 2020s... (not to mention the Clone Wars where if subject is higher than just 1 everyone looks like twins regardless of gender and "nationality" - SDXL is very tunnel-visioned when it comes to different facial texture meshes...)

Then, this model is a bless. Or I simply don't know how to prompt for SDXL, but luckily when I add you on, magic happens and it feels like it translates what I type so that it makes sense.

Course, for horror, ultra-dark (SFW) themes I don't use it, but I think it's not the purpose anyways.

Cheers for making it! ❤

I used this as an enhancer for Crystal Clear XL and after many years of having an image of character in my head featured in my stories...now, I can SEE her!!

The results are truly Beautiful, worth waiting three decades for...I am not kidding, she's been haunting me for a long time. lol

I'll be using the art your enhancer helped create for the Cover Art of the 2nd book in my Cybernoire series, and you'll be getting a due and proper acknowledgement and statement of Thanks at the front of the book, of course.

Thank-you!! My Creation now 'Lives'!!!

Lora works perfectly in Comfyui/with a Rig of 6 gig 1660 super GTX,Ryzen 5 - 2600. 16gigram. Thank you, will upload my results.

This lora is simply mandatory for artists. You get amazing results with this !

I really love this in SDXL is there any version of this or similar in 1.5?

I had avoided this lora for a long time because it just seemed "weird" and I didn't really know what it was for (the intro images produce that impression). However after seeing it had been used to produce the best results for some character loras, I used it where I would typically have a style lora and found it could make loras that were previously not producing good results both artistic and consistent. In particular, I am impressed with how it works with the base style of the lora. Specifically, if the base style of the character lora is illustrated, it will make a more artistic and consistent illustrated style between images. And if the base character lora style is photo realistic, it will produce consistent artistic images in that style. Very happy I started using this.

Excuse me, do you still have an old version of xl_more-art full? Where can I download it

can you shrink this down to 218mb?

Is this one new or just a reupload? I remember this lora since basically sdxl started but here it says it's been uploaded only 2 months ago

i gave this a go. my jaw dropped! bravo.

Очень хорошо эта Lora работает с Pony. У обычной Pony есть прикол с тем, что она может не учитывать многие слова, а добавляя вашу модель всё чинится. Это просто магия какая то. Огромное спасибо.

Creative and beautiful model, you can do real stuff and also sci-fi or fantasy arts without a moment. It's really cool for every usage

Hey LEDADU. XL-MORE-ART-FULL is a great LORA, but now, when used, it generates lots of error messages in COMFYUI 0.12.3 like: "ERROR lora diffusion_model.output_blocks.1.0.in_layers.2.weight unsupported operand type(s) for *: 'NoneType' and 'float'". And because of this, LORA has no effect on generated image. :( Fractal-Art lora, however, works without a problem.