This series is focused on blending cinematic themes with SFW/NSFW. It can generate a range of film, fantasy, realistic, surreal art and fine art. It is great at HDR quality images and cinematic images, with maximum detail.
If you're a fan of images that seem to pop out at you, then this model is for you. This series will always focus on cinematic aesthetics and NSFW ready.
Prompt weight guide below.
Donations to my Whop account will exclusively go towards hardware upgrades. Civitai Buzz will be used for site generation and promotions.
Upgrading my local hardware is my on-going goal so I can train locally.
Local training is imperative today. With Ai data centers there will be a vast drain on resources in the areas they operate. Since model & lora training will no doubt be hosted in the new centers, I want to be able to run locally only. The moment my Ai hobby causes actual harm, on any scale, is when I start collecting stamps instead.
General Prompt Guide
Weights in my prompts are primarily for testing the model's capabilities and trying to uncover limitations during test generation. However, I've created the tutorial below to help understand how model weights can be used more effectively.
Positive Attention
Weights between 0.5 & 2.0 are usually sufficient.
I sometimes use weights between 2.0 & 3.0 when I want to force attention to a specific ethnicity for testing.
Higher weights are also helpful when you want to magnify size or specific details. The difference between: large breasts, (large breasts:1.2), & (large breasts:1.8) is obvious.
Recognized weights are usually -5.0 to 5.0 with Dynamic Prompts.
Attention Equation
The parentheses ( ) on their own adds attention weight of 1.1. Unless you specify a weight.
(hyper-realistic) = 1.1
((hyper-realistic)) = 1.21
((hyper-realistic:1.3)) = 1.43
Brackets [ ] are negative attention weight of 0.91, not negative weight, it's just attention that starts under the 1.0 default of all words without weights. Stacking them decreases that number.
[deep shadows] = 0.91
[[deep shadows]] = 0.81
[[[deep shadows]]] = 0.72
Negative Attention
Weights over 3.5 are not usually effective or are a detriment during tokenization.
Negative weights do not have much effect in prompts.
Using too many high weights in the same prompt can cause distortions & deformities.
Using high weights will give specific keywords more attention, but that is often at the expense of other keywords.
Bonus
Prompt with high weights:
(HDR:2.2), cinematic, (natural sunlight:1.23), (hyper-realistic:1.45), (full body shot:0.6), masterpiece, incredible details, [feeling of solitude], mysterious, (deep shadows:0.72), Beautiful (30yo Malaysian woman:2.52), hazel hair, (mid_length grungy hairstyle :1.43), (natural large breasts:1.6), (wearing an unbuttoned stylish silk shirt), (brown pearl g-string, leather lace sandals:1.41), (sexy pose, reclining slightly on elbow:1.68) on a stunning designer wicker couch, at a rustic beach house rooftop with driftwood furniture, seashell decor, a view of the ocean, perfect for casual summer parties, sensual atmosphere, beautiful, guilty pleasure, shot on Canon EOS 5D, RAW.
Prompt without high weights:
(HDR:0.86), cinematic, (natural sunlight:0.8), (hyper-realistic), (full body shot:0.6), masterpiece, incredible details, feeling of solitude, mysterious, (deep shadows:0.72), Beautiful (30yo Malaysian woman:1.32), hazel hair, (mid_length grungy hairstyle), (natural large breasts), (wearing an unbuttoned stylish silk shirt), (brown pearl g-string, leather lace sandals), (sexy pose, reclining slightly on elbow:1.12) on a stunning designer wicker couch, at a rustic beach house rooftop with driftwood furniture, seashell decor, a view of the ocean, perfect for casual summer parties, sensual atmosphere, beautiful, guilty pleasure, shot on Canon EOS 5D, RAW.
Description
Settings I use:
DPM SDE
30 Steps
4.6 CFG
FAQ
Comments (9)
I have the same opinion than for Reality Bound, the model can generate amazing images, but prompt understanding and adherence need improvement.
Think v2.0 is better than v3.0 in this regard.
I hear you. Working on a new process to address this without harming other traits.
Regarding 2.0, while it can produce some really high quality output, it's prompt adherence is one of the worst I've ever seen. I've never seen a model where you need to use (1.6) weights to get it to slightly grasp at a concept. Probably won't try 3.0 for this reason. A model should not need to use extra emphasis on anything except in rare cases, and some of your base prompts are using :2, I've never even seen that before. Any other model using prompts like this would cause it to explode.
Try merging it with other models that you like that might be missing the creativity you want from this one.
My prompts are generated from wildcards. They all have emphasis by default. It's how I test models to see what they can or can't do. I would rather it fail adherence on some prompts than give me a cursed or distorted image.
You're complaining about my prompts as if that's some sort of carefully scaled measure of what it needs to work. No, of course not. Also, if :2, "blows up" your other models, then that is why I test that way. To make sure it interprets the prompt rather than give up on trying.
The issue is, I can't get it to do literally anything I prompt it for, unless I use lots and lots of emphasis... and even then, it struggles. I've been using these models a long time and know exactly how to prompt, and how each base model prompts differently. I'm not trying anything super creative. Just normal base prompts for testing.
@Mohzaic I apologize for being condescending. I added the prompt tutorial because there were other related comments...and I like writing them. Would you mind sharing a prompt example, or several, so I can use them for testing models? Privately, so we can chat about it, if that's ok.
Also, have you tried 4.0? I would like to know what improvements need to be made. My prompts may not be showing problem areas others may have.
@ironically_irascible I'll give 4.0 a try and get back to you.