-
Notifications
You must be signed in to change notification settings - Fork 26
Advanced config
The config.properties file is a configuration file that allows users to customize the behavior and settings of the SDXL-DiscordBot. This wiki provides a detailed breakdown of each section in the file.
This section is for configurations related to text-to-image generation when running the bot locally using ComfyUI.
- CONFIG: Path to the JSON configuration file for text-to-image generation.
- PROMPT_NODES: Comma-separated list of node IDs for positive text prompts.
- NEG_PROMPT_NODES: Comma-separated list of node IDs for negative text prompts.
- RAND_SEED_NODES: Comma-separated list of node IDs where random seeds will be injected.
This section is for configurations related to image-to-image transformations when running the bot locally using ComfyUI.
- CONFIG: Path to the JSON configuration file for image-to-image transformations.
- PROMPT_NODES: Node ID for the primary image input.
- NEG_PROMPT_NODES: Node ID for the negative image input.
- RAND_SEED_NODES: Node ID where a random seed will be injected for variations.
- FILE_INPUT_NODES: Node ID for file-based image inputs.
This section is for configurations related to image upscaling when running the bot locally using ComfyUI.
- CONFIG: Path to the JSON configuration file for image upscaling.
- FILE_INPUT_NODES: Node ID for file-based image inputs.
This section is for configurations related to text-to-image generation when using the Stability AI API.
- ENGINE: The specific engine version to use for the text-to-image generation.
- CFG: Configuration scale for the model.
- WIDTH: Desired width of the generated image.
- HEIGHT: Desired height of the generated image.
- SAMPLES: Number of sample images to generate.
- SAMPLER: The sampling method to use.
- STEPS: Number of steps for the generation process.
This section is for configurations related to image-to-image transformations when using the Stability AI API.
- ENGINE: The specific engine version to use for the image-to-image transformation.
- INIT_IMAGE_MODE: The mode to initialize the image.
- IMAGE_STRENGTH: Strength of the initial image.
- CFG: Configuration scale for the model.
- SAMPLES: Number of sample images to generate.
- SAMPLER: The sampling method to use.
- STEPS: Number of steps for the transformation process.
This section is for configurations related to image upscaling when using the Stability AI API.
- ENGINE: The specific engine version to use for the image upscaling.
- WIDTH: Desired width of the upscaled image.
- SEED: Seed value for the upscaling process.
- STEPS: Number of steps for the upscaling process.
- CFG: Configuration scale for the model.
If you're looking to customize SDXL-DiscordBot further with your own ComfyUI workflows, you're at the right place. Follow this step-by-step guide to seamlessly integrate custom workflows:
Begin by designing your desired workflow in ComfyUI. Ensure you have a clear vision of what you want to achieve with this custom workflow.
When setting up your workflow, make sure the output images have the prefix "final_output". This ensures the bot can recognize and process these images correctly.
- Navigate to the developer options within ComfyUI.
- Look for the option to save your workflow.
- Choose the "API Format" option. This will save your workflow as a JSON file, which is essential for the next steps.
Take the JSON file you just saved and upload it to the comfyUI-workflows folder designated for the bot.
Open the bot's properties file. You'll need to update the CONFIG location to point to the new workflow JSON you've just uploaded.
In the same properties file:
- Identify and specify the Prompt nodes.
- Specify the Negative Prompt nodes.
- Point out the Random Seed nodes.
- Finally, identify the File Input nodes.
This step ensures the bot knows where to pull specific data from within your custom workflow.
Once all the above steps are completed, restart the bot. Post-restart, your custom ComfyUI workflow should be integrated and operational.