Welcome to Synthetic Forge

Preamble

What is Synthetic Forge AI for?

Synthetic Forge AI is a tool programmed to help AI content creators automate their content production. Version Alpha 0.1.0 allows you, through the creation of configuration files, to obtain several different prompts without having to write them constantly by hand.

Configuration of Synthetic Forge AI

Prerequisites

• Have Stable Diffusion WebUI installed.
• Add the parameters "--api --listen" for the "set COMMANDLINE_ARGS=" option in the "webui-user.bat" file.
• Launch the program and select the model you want to use.

File "config/config.json"

To control the automatic generation of your content, the "config.json" file available in the "config" folder must be modified according to the needs of your model and your intention. Each field must be filled in to make the application work.

Explanation of Fields

URL

This field should be filled with the URL returned by Stable Diffusion WebUI when launched. By default, its value is "http://127.0.0.1:7860" or "http://localhost:7860". Modify this field according to your needs.

Token

Your access key to the Synthetic Forge AI Prompt Generation software. If you don't have one, contact me on Patreon (https://www.patreon.com/SyntheticWAIfu) or X/Twitter (https://twitter.com/SynthWaifu_AI).

Counter

This field defines the number of prompts that Synthetic Forge AI should generate automatically. All prompts will be saved after generation in the "outputs/logs/" folder.

Data

Information sent to the Stable Diffusion API, contains the following fields:

• Prompt: Should remain empty (configured later).
• Styles: If you want to apply specific styles to all your prompts, fill this field with the names of the styles you want, separate each style by a comma. Example: ["style1", "style2"].
• Negative: Should remain empty (configured later).
• Sampler Name: Should remain empty (configured later).
• N_Iter: Number of images you want for each generated prompt.
• Steps: Number of steps you want for each generated prompt, must be defined according to your model.
• Cfg_scale: Strength of the generated prompt, must be defined according to your model (Higher values result in the model following the prompt more closely).
• Width: Width of the generated image.
• Height: Height of the generated image.
• Denoising_Strength: Strength of denoising of the generated image, must be defined according to your model and Sampler. Used to reduce anomalies in the generated image.
• Seed: Image generation seed, usually set to -1 for random generation.
• Enable_Hr: If you want to enable high resolution for your prompts, set this field to "true", otherwise to "false".
• Hr_Upscale: If you have enabled high resolution, set the image upscaling factor, often "1.5" or "2".
• Hr_Upscaler: Sampler used for image upscaling, must be defined according to your model.
• Hr_Second_Pass_Steps: Number of steps for the second pass of image upscaling, must be defined according to your model.
• Save_Images: If you want to save the generated images in the folders defined in the Stable Diffusion WebUI configuration, set this field to "true", otherwise to "false".

Presets_Style

If you want to apply specific styles to all your prompts, set this field to "true", otherwise to "false". If set to "true", fill in the "Styles" field in the "Data" field.

Presets_Outfits

If you want to apply specific outfits to all your prompts, set this field to "true", otherwise to "false", similarly to the "Presets_Style" field, apply your styles in the "Styles" field in the "Data" field.

Is_A_Girl

If you want to generate girl prompts, set this field to "true", otherwise "false", this will apply "1girl" to each generated prompt.

Samplers

List of Samplers you want to use for generating your prompts, separate each Sampler by a comma. Example: ["sampler1", "sampler2"].

Sampler_Index

Index of the Sampler you want to use for generating your prompts compared to your Sampler list, the index starts at 0. In our example, if you want to use Sampler 2, put "1".

Hires_Fixes

List of Upscalers you want to use for generating your prompts, separate each Upscaler by a comma. Example: ["upscaler1", "upscaler2"].

Hires_Fixes_Index

Index of the Upscaler you want to use for generating your prompts compared to your list of Upscalers, the index starts at 0. In our example, if you want to use Upscaler 2, put "1".

Negatives

List of Negatives you want to use for generating your prompts, separate each Negative by a comma. Example: ["negative1", "negative2"].

Negatives_Index

Index of the Negative you want to use for generating your prompts compared to your list of Negatives, the index starts at 0. In our example, if you want to use Negative 2, put "1".

Presets

Object comprising several fields

• Accessories: Path to the file listing the accessories available randomly for each generated prompt. In the file, list the accessories you want to use, separate each accessory by a comma. Example: accessory1, accessory2
• Base: Path to the file listing the base prompt you want to use for each generated prompt. In the file, list the prompts you want to use.
• Body_Type: Path to the file listing the body types available randomly for each generated prompt. In the file, list the body types you want to use, separate each body type by a comma. Example: type1, type2
• Bottom_Outfits: Path to the file listing the bottoms available randomly for each generated prompt. In the file, list the bottoms you want to use, separate each bottom by a comma. Example: bottom1, bottom2
• Breast_Type: Path to the file listing the breast types available randomly for each generated prompt. In the file, list the breast types you want to use, separate each breast type by a comma. Example: type1, type2
• Character_Poses: Path to the file listing the poses available randomly for each generated prompt. In the file, list the poses you want to use, separate each pose by a comma. Example: pose1, pose2
• Facial_Expressions: Path to the file listing the facial expressions available randomly for each generated prompt. In the file, list the facial expressions you want to use, separate each facial expression by a comma. Example: expression1, expression2
• Hair_Colors: Path to the file listing the hair colors available randomly for each generated prompt. In the file, list the hair colors you want to use, separate each hair color by a comma. Example: color1, color2
• Hair_Styles: Path to the file listing the hair styles available randomly for each generated prompt. In the file, list the hair styles you want to use, separate each hair style by a comma. Example: style1, style2
• Hair_Accessories: Path to the file listing the hair accessories available randomly for each generated prompt. In the file, list the hair accessories you want to use, separate each hair accessory by a comma. Example: accessory1, accessory2
• Hair_Lengths: Path to the file listing the hair lengths available randomly for each generated prompt. In the file, list the hair lengths you want to use, separate each hair length by a comma. Example: length1, length2
• Locations: Path to the file listing the locations available randomly for each generated prompt. In the file, list the locations you want to use, separate each location by a comma. Example: location1, location2
• Photo_Types: Path to the file listing the photo types available randomly for each generated prompt. In the file, list the photo types you want to use, separate each photo type by a comma. Example: type1, type2
• Shoes_Feet: Path to the file listing the shoes available randomly for each generated prompt. In the file, list the shoes you want to use, separate each shoe by a comma. Example: shoe1, shoe2
• DayTimes: Path to the file listing the times of day available randomly for each generated prompt. In the file, list the times of day you want to use, separate each time of day by a comma. Example: time1, time2
• Top_Outfits: Path to the file listing the tops available randomly for each generated prompt. In the file, list the tops you want to use, separate each top by a comma. Example: top1, top2
• Weather: Path to the file listing the types of weather available randomly for each generated prompt. In the file, list the types of weather you want to use, separate each type of weather by a comma. Example: weather1, weather2

Once filled, launch the program and wait for the prompts to be generated. The prompts will be saved in the "outputs/logs/" folder. All your images will be saved in the folder defined in the Stable Diffusion WebUI configuration.