Description
The fofr/sdxl-emoji tool is an AI model that was refined using Apple Emojis. It can be accessed via an API on the Replicate platform.
This tool enables users to produce and modify images using prompts and parameters. Users are able to input a TOK emoji of a man, and add a negative prompt to customize the output.
Also, users have the option to upload images for use within the model or to provide a mask to specify image regions to preserve or inpaint.
They can define the width, height, and count of output images. The tool provides options for solvers and scheduling to denoise and refine images, including Multistep, Heun, Discrete Karras, DPM, K_EULER_ANCESTRAL, and K_EULER.
Users can adjust the guidance scale and set a prompt strength for the img2img and inpaint functions. The tool features customization of the random seed. Further capabilities include the ability to choose a refine style, manage the level of noise with the expert ensemble refiner, and adjust refinement steps when using the base_image_refiner.
Users may also apply a watermark to their generated images. The model outputs results in approximately 15.18 seconds and allows users to download the generated image or check out examples.
Predictions run on Nvidia A40 (Large) GPU hardware, and the tool is hosted on the Replicate platform.
