VINTAGE

This AI tool streamlines the input parameters to their most fundamental form for casual users inexperienced with AI, optimizing the settings to maximize the value of your credits. This tool was made for entertainment.Learn more of this Ai Tool here: https://tensor.art/articles/785838505587753778Instructions:Select a Generation Seed: You don't need to alter the remaining seed values for your initial image generation.Choose an anime reference to guide your prompt to that style; this is useful if your prompt is straightforward.Enter your preferred prompt in the tidytag format (the way we prompt in PonyV6) in this field. Natural language is not advised at all.In the 'face detailer' area, type a prompt to customize the facial aspect.Choose a 'Resolution': In PonyV6, resolution is essential to achieving the result you want; various outputs perform better at particular resolutions.Input Prompt tips and tricks:Use the 'Variance seed' and 'ESND' to get variations of specific outputs, remember keeping the 'Generation (seed)' static.Make sure to have a correct tokenization employing the best tags (view checkpoint model and LoRAs to know more). Danbooru tags tends to move the output into NSFW territory.The prompt in the 'face detailer' field won't affect your input prompt, it won't contribute for its tokenization, this prompt is independent of your input prompt.Syntax like 'WAS A1111', 'AND', 'BREAK' are invalid, please do not try it,ADetailer (facedetailer): the tool has been limited to recognize up to 5 faces at the same time.MIscellaneous📘 What is the 75 prompt limit in stable diffusion?🦄 Tokenization in Ponyv6The way tokens work is that they are batched in groups of 75. If you have <=75 tokens, the diffusion process has only one batch of tokens. If you have 78 tokens, you will have one batch of 75 tokens and one batch of 3 tokens, and the attention of the diffusion process will be divided in half. This will give you less of your first 75 tokens and may overemphasize the 3 prompt tokens in the second batch. The more batches of tokens, the less attention on each batch.🦄 Pony Diffusion XL Recognized Characters List🦄 Danbooru Tagging Visualization for PonyXL/AutismMix⚙️ Preloaded ItemsYou don't need to provide score tags like score_9, score_8_up, score_7_up, and source_anime because they are already proloaded.Although they are not included, rating tags like rating_safe, rating_questionable, and rating_explicit can be added.Although they are not included, source tags like source_furry and source_pony can be added.Negative prompt: not accessible since it has already been preloaded.V2.1 LOG151124 | V2.1 (sailor moon wildcards added)111424 | NSFW is supported, (make sure to prompt the correct keywords).V1.0 Halloween Edition LOG111424 | No longer availible.✉️ Discord Thread: https://discord.com/channels/1108668896959025214/1297239228010467479

Ever wonder how you would have looked back in those Vintage Days? Wonder no more - just drop a clear face shot into the input and enjoy some random vintageness courtesy of @PictureT's awesome lora - Western Styles

Prompt:If you are having problems with pics coming out weird:change the prompt in that specific part - make the prompt more clear(Ex: instead of "his fingers wrapping around the sword" use "he holds the sword") don't use one super large sentences to describe one thing - break it into pieces (Ex: instead of "a beautiful woman with red long hair wears a green old cape and sits in a medieval castle" use "a beautiful woman with red long hair, wearing a green cape, sitting in a medieval castle")don't break the flow of the prompt - write it like a sequence(Ex: instead of "a beautiful woman, wearing a green cape, sitting in a medieval castle, with red long hair" use "a beautiful woman with red long hair, wearing a green cape, sitting in a medieval castle")the "abstract prompt" (image to prompt) has a weird natural language that messes up almost every LoRa I tried - correct its language and order of description

Go back in time and generate a photo of you or someone else.