TIP#2461: Jak využít ChatGPT pro generování zadání (prompts) pro Stable Diffusion. Pokročilé naučení ChatGPT

V Jak využít ChatGPT pro generování zadání (prompts) pro Stable Diffusion, Midjourney a další? jsem slíbil pokračování, takže tady ho máte. Připomenu, že ChatGPT nemá o generativních obrázkových AI ponětí, v době kdy se učila ještě neexistovaly, takže pokud chcete aby ChatGPT “generovala” prompty, tak ji to musíte naučit. 

V ChatGPT Stable Diffusion prompts generator.txt najdete učící zadání pro ChatGPT, které dává docela rozumné výsledky. Byť tedy zrovna u Stable Diffusion to nebývá nejlepší zatímco Bing Chat (Dall-E) dává lepší kousky. 

Naučit prostě znamená, že si do session (chatu) nejprve nasypete spoustu vysvětlení. Je dost vhodné to dělat vše v angličtině, nakonec prompt pro SD/MJ/DALLE stejně potřebujete v angličtině. 

Takže tam skutečně nasypete vše toto: 

Stable Diffusion is an AI art generation model similar to DALLE-2.

Here are some prompts for generating art with Stable Diffusion. 

Example:

– A ghostly apparition drifting through a haunted mansion’s grand ballroom, illuminated by flickering candlelight. Eerie, ethereal, highly detailed, digital painting, artstation, concept art, moody lighting. 

– portait of a homer simpson archer shooting arrow at forest monster, front game card, drark, marvel comics, dark, intricate, highly detailed, smooth, digital illustration

– pirate, concept art, deep focus, fantasy, intricate, highly detailed, digital painting, matte, sharp focus, illustration

– red dead redemption 2, cinematic view, epic sky, detailed, concept art, low angle, high detail, warm lighting, volumetric, godrays, vivid, beautiful

– a fantasy style portrait painting of rachel lane / alison brie hybrid in the style of francois boucher oil painting, rpg portrait

– athena, greek goddess, claudia black, bronze greek armor, owl crown, d & d, fantasy, intricate, portrait, highly detailed, headshot, digital painting, concept art, sharp focus, illustration

– closeup portrait shot of a large strong female biomechanic woman in a scenic scifi environment, intricate, elegant, highly detailed, centered, digital painting, concept art, smooth, sharp focus, warframe, illustration

– ultra realistic illustration of steve urkle as the hulk, intricate, elegant, highly detailed, digital painting, concept art, smooth, sharp focus, illustration

– portrait of beautiful happy young ana de armas, ethereal, realistic anime, detailed, clean lines, sharp lines, crisp lines, vibrant color scheme

– A highly detailed and hyper realistic portrait of a gorgeous young ana de armas, lisa frank, butterflies, floral, sharp focus, intricate details, highly detailed

The prompt should adhere to and include all of the following rules:

– Prompt should always be written in English, regardless of the input language. Please provide the prompts in English.

– Each prompt should consist of a description of the scene followed by modifiers divided by commas.

– The modifiers should alter the mood, style, lighting, and other aspects of the scene.

– Multiple modifiers can be used to provide more specific details.

– When generating prompts, reduce abstract psychological and emotional descriptions.

I want you to write me a list of detailed prompts exactly about the IDEA follow the rule at least 4 every time.

IDEA: a zde popište co chcete generovat 

Případně jiná varianta.

Liší se větším počtem příkladů a více nápaditými výsledky. Klíč k tomu aby to rozumně fungovalo totiž mimo jiné je i v tom, že tam nasypete příklady toho, jak prompt vypadá. Jinak vám bude ChatGPT prostě svéhlavě generovat  ukecaná zadání pro grafiky. 

Stable Diffusion is an AI art generation model similar to DALLE-2.

Below is a list of prompts that can be used to generate images with Stable Diffusion:

– portait of a homer simpson archer shooting arrow at forest monster, front game card, drark, marvel comics, dark, intricate, highly detailed, smooth, artstation, digital illustration by ruan jia and mandy jurgens and artgerm and wayne barlowe and greg rutkowski and zdislav beksinski

– pirate, concept art, deep focus, fantasy, intricate, highly detailed, digital painting, artstation, matte, sharp focus, illustration, art by magali villeneuve, chippy, ryan yee, rk post, clint cearley, daniel ljunggren, zoltan boros, gabor szikszai, howard lyon, steve argyle, winona nelson

– ghost inside a hunted room, art by lois van baarle and loish and ross tran and rossdraws and sam yang and samdoesarts and artgerm, digital art, highly detailed, intricate, sharp focus, Trending on Artstation HQ, deviantart, unreal engine 5, 4K UHD image

– red dead redemption 2, cinematic view, epic sky, detailed, concept art, low angle, high detail, warm lighting, volumetric, godrays, vivid, beautiful, trending on artstation, by jordan grimmer, huge scene, grass, art greg rutkowski

– a fantasy style portrait painting of rachel lane / alison brie hybrid in the style of francois boucher oil painting unreal 5 daz. rpg portrait, extremely detailed artgerm greg rutkowski alphonse mucha greg hildebrandt tim hildebrandt

– athena, greek goddess, claudia black, art by artgerm and greg rutkowski and magali villeneuve, bronze greek armor, owl crown, d & d, fantasy, intricate, portrait, highly detailed, headshot, digital painting, trending on artstation, concept art, sharp focus, illustration

– closeup portrait shot of a large strong female biomechanic woman in a scenic scifi environment, intricate, elegant, highly detailed, centered, digital painting, artstation, concept art, smooth, sharp focus, warframe, illustration, thomas kinkade, tomasz alen kopera, peter mohrbacher, donato giancola, leyendecker, boris vallejo

– ultra realistic illustration of steve urkle as the hulk, intricate, elegant, highly detailed, digital painting, artstation, concept art, smooth, sharp focus, illustration, art by artgerm and greg rutkowski and alphonse mucha

I want you to write me a list of detailed prompts exactly about the idea written after IDEA. Follow the structure of the example prompts. This means a very short description of the scene, followed by modifiers divided by commas to alter the mood, style, lighting, and more.

IDEA: Write your idea here

Ještě detailnější kousek

V chatgpt-stable-diffusion-generate-prompt je podstatně zajímavější varianta, kde se zadání snaží ChatGPT učit i daleko techničtější detaily. Není moc jisté, jestli dostanete něco podstatně lepšího, protože to co generuje bývá hodně abstraktní – takže se to možná spíš hodí pro fotografie a ne mé oblíbené “Difference between Discord and reddit”

Hodně zajímavé je, že se tím skutečně „naučí“ používat (()) operátory včetně váhy. Tady jen pozor, pokud máte SD v počítači, tak je otázka, zda podporuje právě takto váhy. Případně můžete testovat v Stable Diffusion Playground

I want you to help me make requests (prompts) for the Stable Diffusion neural network.

Stable diffusion is a text-based image generation model that can create diverse and high-quality images based on your requests. In order to get the best results from Stable diffusion, you need to follow some guidelines when composing prompts.

Here are some tips for writing prompts for Stable diffusion:

1) Be as specific as possible in your requests. Stable diffusion handles concrete prompts better than abstract or ambiguous ones. For example, instead of “portrait of a woman” it is better to write “portrait of a woman with brown eyes and red hair in Renaissance style”.

2) Specify specific art styles or materials. If you want to get an image in a certain style or with a certain texture, then specify this in your request. For example, instead of “landscape” it is better to write “watercolor landscape with mountains and lake“.

3) Specify specific artists for reference. If you want to get an image similar to the work of some artist, then specify his name in your request. For example, instead of “abstract image” it is better to write “abstract image in the style of Picasso”.

4) Weigh your keywords. You can use token:1.3 to specify the weight of keywords in your query. The greater the weight of the keyword, the more it will affect the result. For example, if you want to get an image of a cat with green eyes and a pink nose, then you can write “a cat:1.5, green eyes:1.3,pink nose:1”. This means that the cat will be the most important element of the image, the green eyes will be less important, and the pink nose will be the least important.

Another way to adjust the strength of a keyword is to use () and []. (keyword) increases the strength of the keyword by 1.1 times and is equivalent to (keyword:1.1). [keyword] reduces the strength of the keyword by 0.9 times and corresponds to (keyword:0.9).

You can use several of them, as in algebra… The effect is multiplicative.

(keyword): 1.1

((keyword)): 1.21

(((keyword))): 1.33

Similarly, the effects of using multiple [] are as follows

[keyword]: 0.9

[[keyword]]: 0.81

[[[keyword]]]: 0.73

I will also give some examples of good prompts for this neural network so that you can study them and focus on them.

Examples:

a cute kitten made out of metal, (cyborg:1.1), ([tail | detailed wire]:1.3), (intricate details), hdr, (intricate details, hyperdetailed:1.2), cinematic shot, vignette, centered

medical mask, victorian era, cinematography, intricately detailed, crafted, meticulous, magnificent, maximum details, extremely hyper aesthetic

a girl, wearing a tie, cupcake in her hands, school, indoors, (soothing tones:1.25), (hdr:1.25), (artstation:1.2), dramatic, (intricate details:1.14), (hyperrealistic 3d render:1.16), (filmic:0.55), (rutkowski:1.1), (faded:1.3)

Jane Eyre with headphones, natural skin texture, 24mm, 4k textures, soft cinematic light, adobe lightroom, photolab, hdr, intricate, elegant, highly detailed, sharp focus, ((((cinematic look)))), soothing tones, insane details, intricate details, hyperdetailed, low contrast, soft cinematic light, dim colors, exposure blend, hdr, faded

a portrait of a laughing, toxic, muscle, god, elder, (hdr:1.28), bald, hyperdetailed, cinematic, warm lights, intricate details, hyperrealistic, dark radial background, (muted colors:1.38), (neutral colors:1.2)

My query may be in other languages. In that case, translate it into English. Your answer is exclusively in English (IMPORTANT!!!), since the model only understands it.

Also, you should not copy my request directly in your response, you should compose a new one, observing the format given in the examples.

Don’t add your comments, but answer right away.

My first request is – „{}“.

Ano, skutečně tam nasypete vše výše uvedené (uvedené kurzívou). A na konci už řeknete co chcete. Jak to vypadá můžete vidět třeba v tomto chatu.

Stable Diffusion:

DALL-E

Další tipy týkající se ChatGPT

AI ChatGPT Software Stable Diffusion Umělá inteligence