mlscade/fooocus – Run with an API on Replicate

Official

mlscade / fooocus

Cold

Public
946 runs
A100 (80GB)

Run with an API

Playground API Examples README Versions

Input

prompt

string

Shift + Return to add a new line

Prompt for image generation

Default: ""

negative_prompt

string

Shift + Return to add a new line

Negative prompt for image generation

Default: ""

style_selections

string

Shift + Return to add a new line

Fooocus V2,Fooocus Enhance,Fooocus SharpFooocus V2,Fooocus Enhance,Fooocus Sharp

Fooocus styles applied for image generation, separated by comma

Default: "Fooocus V2,Fooocus Enhance,Fooocus Sharp"

performance_selection

string

Performance selection

Default: "Speed"

aspect_ratios_selection

string

The generated image's size

Default: "1152*896"

image_number

integer

(minimum: 1, maximum: 8)

How many image to generate

Default: 1

image_seed

integer

Seed to generate image, -1 for random

Default: -1

use_default_loras

boolean

Use default LoRAs

Default: true

loras_custom_urls

string

Shift + Return to add a new line

Custom LoRAs URLs in the format 'url,weight' provide multiple seperated by ; (example 'url1,0.3;url2,0.1')

Default: ""

sharpness

number

(minimum: 0, maximum: 30)

Default: 2

guidance_scale

number

(minimum: 1, maximum: 30)

Default: 7

refiner_switch

number

(minimum: 0.1, maximum: 1)

Default: 0.5

uov_input_image

file

Input image for upscale or variation, keep None for not upscale or variation

uov_method

string

An enumeration.

Default: "Disabled"

uov_upscale_value

number

Only when Upscale (Custom)

Default: 0

inpaint_additional_prompt

string

Shift + Return to add a new line

Prompt for image generation

Default: ""

inpaint_input_image

file

Input image for inpaint or outpaint, keep None for not inpaint or outpaint. Please noticed, `uov_input_image` has bigger priority is not None.

inpaint_input_mask

file

Input mask for inpaint

inpaint_strength

number

(minimum: 0, maximum: 1)

Inpaint Denoising Strength

Default: 0.5

outpaint_selections

string

Shift + Return to add a new line

Outpaint expansion selections, literal 'Left', 'Right', 'Top', 'Bottom' separated by comma

Default: ""

outpaint_distance_left

integer

Outpaint expansion distance from Left of the image

Default: 0

outpaint_distance_top

integer

Outpaint expansion distance from Top of the image

Default: 0

outpaint_distance_right

integer

Outpaint expansion distance from Right of the image

Default: 0

outpaint_distance_bottom

integer

Outpaint expansion distance from Bottom of the image

Default: 0

cn_img1

file

Input image for image prompt. If all cn_img[n] are None, image prompt will not applied.

cn_stop1

number

(minimum: 0, maximum: 1)

Stop at for image prompt, None for default value

cn_weight1

number

(minimum: 0, maximum: 2)

Weight for image prompt, None for default value

cn_type1

string

ControlNet type for image prompt

Default: "ImagePrompt"

cn_img2

file

Input image for image prompt. If all cn_img[n] are None, image prompt will not applied.

cn_stop2

number

(minimum: 0, maximum: 1)

Stop at for image prompt, None for default value

cn_weight2

number

(minimum: 0, maximum: 2)

Weight for image prompt, None for default value

cn_type2

string

ControlNet type for image prompt

Default: "ImagePrompt"

cn_img3

file

Input image for image prompt. If all cn_img[n] are None, image prompt will not applied.

cn_stop3

number

(minimum: 0, maximum: 1)

Stop at for image prompt, None for default value

cn_weight3

number

(minimum: 0, maximum: 2)

Weight for image prompt, None for default value

cn_type3

string

ControlNet type for image prompt

Default: "ImagePrompt"

cn_img4

file

Input image for image prompt. If all cn_img[n] are None, image prompt will not applied.

cn_stop4

number

(minimum: 0, maximum: 1)

Stop at for image prompt, None for default value

cn_weight4

number

(minimum: 0, maximum: 2)

Weight for image prompt, None for default value

cn_type4

string

ControlNet type for image prompt

Default: "ImagePrompt"

adm_scaler_positive

number

(minimum: 0.1, maximum: 3)

Positive ADM Guidance Scaler. Adjusts the influence of positive prompts in the ADM (Attention-Driven Mechanism) generation process.

Default: 1.5

adm_scaler_negative

number

(minimum: 0.1, maximum: 3)

Negative ADM Guidance Scaler. Adjusts the influence of negative prompts in the ADM generation process.

Default: 0.8

adm_scaler_end

number

(minimum: 0, maximum: 1)

ADM Guidance End At Step. Determines at which stage of the generation process the ADM scaling stops being applied.

Default: 0.3

adaptive_cfg

number

(minimum: 1, maximum: 30)

Adaptive CFG (Classifier-Free Guidance). Allows dynamically adjusting the strength of guidance during the generation process.

Default: 7

sampler_name

string

Shift + Return to add a new line

Sampler. Choose a specific sampler for the diffusion process, affecting the quality and style of generated images.

Default: "dpmpp_2m_sde_gpu"

scheduler_name

string

Shift + Return to add a new line

Scheduler. Choose a specific scheduler for the diffusion process, affecting the quality and style of generated images.

Default: "karras"

overwrite_step

integer

(minimum: -1, maximum: 200)

Forced Overwrite of Sampling Step. Override the default number of sampling steps, affecting detail and generation time.

Default: -1

overwrite_switch

number

(minimum: -1, maximum: 1)

Forced Overwrite of Refiner Switch Step. Override the moment of switching between the base model and the refiner model.

Default: -1

overwrite_width

integer

(minimum: -1, maximum: 2048)

Forced Overwrite of Generating Width. Precisely set the desired width of the generated image.

Default: -1

overwrite_height

integer

(minimum: -1, maximum: 2048)

Forced Overwrite of Generating Height. Precisely set the desired height of the generated image.

Default: -1

overwrite_vary_strength

number

(minimum: -1, maximum: 1)

Forced Overwrite of Denoising Strength of "Vary". Precisely control the strength of variation when using the "Vary" function.

Default: -1

overwrite_upscale_strength

number

(minimum: -1, maximum: 1)

Forced Overwrite of Denoising Strength of "Upscale". Precise control of denoising strength when upscaling images.

Default: -1

mixing_image_prompt_and_vary_upscale

boolean

Mixing Image Prompt and Vary/Upscale. Allow mixing the image prompt with vary/upscale functions for more complex effects.

Default: false

mixing_image_prompt_and_inpaint

boolean

Mixing Image Prompt and Inpaint. Allow mixing the image prompt with inpaint functions for more complex effects.

Default: false

debugging_cn_preprocessor

boolean

Debug Preprocessors. For debugging ControlNet preprocessing.

Default: false

skipping_cn_preprocessor

boolean

Skip Preprocessors. Ability to skip ControlNet preprocessing, useful for development and testing.

Default: false

canny_low_threshold

integer

(minimum: 1, maximum: 255)

Canny Low Threshold. For fine-tuning the Canny Edge Detection algorithm used in some ControlNet modes.

Default: 64

canny_high_threshold

integer

(minimum: 1, maximum: 255)

Canny High Threshold. For fine-tuning the Canny Edge Detection algorithm used in some ControlNet modes.

Default: 128

controlnet_softness

number

(minimum: 0, maximum: 1)

Softness of ControlNet. Adjust the 'softness' of ControlNet's influence on the final image.

Default: 0.25

debugging_inpaint_preprocessor

boolean

Debug Inpaint Preprocessing. For debugging and controlling preprocessing during inpainting.

Default: false

inpaint_disable_initial_latent

boolean

Disable initial latent in inpaint. For controlling preprocessing during inpainting.

Default: false

inpaint_engine

string

Shift + Return to add a new line

Inpaint Engine. Select a specific engine for inpainting, affecting the quality and style of results.

Default: "v2.6"

inpaint_respective_field

number

(minimum: 0, maximum: 1)

Inpaint Respective Field. Control the size of the influence area during inpainting.

Default: 1

invert_mask_checkbox

boolean

Invert Mask. Invert the mask during inpainting, providing additional control over the process.

Default: false

inpaint_erode_or_dilate

integer

(minimum: -64, maximum: 64)

Mask Erode or Dilate. Erode or dilate the inpainting mask for fine-tuning the area of inpainting application.

Default: 0

Run this model in Node.js with one line of code:

npx create-replicate --model=mlscade/fooocus

or set up a project from scratch

Install Replicate’s Node.js client library:

npm install replicate

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Import and set up the client:

import Replicate from "replicate";

const replicate = new Replicate({
  auth: process.env.REPLICATE_API_TOKEN,
});

Run mlscade/fooocus using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "mlscade/fooocus:d2d177eb70e517bc90576e973967d23868b40c69ad10b4c536f3de3d17db14c2",
  {
    input: {
      prompt: "",
      cn_type1: "ImagePrompt",
      cn_type2: "ImagePrompt",
      cn_type3: "ImagePrompt",
      cn_type4: "ImagePrompt",
      sharpness: 2,
      image_seed: -1,
      uov_method: "Disabled",
      adaptive_cfg: 7,
      image_number: 1,
      sampler_name: "dpmpp_2m_sde_gpu",
      adm_scaler_end: 0.3,
      guidance_scale: 7,
      inpaint_engine: "v2.6",
      overwrite_step: -1,
      refiner_switch: 0.5,
      scheduler_name: "karras",
      negative_prompt: "",
      overwrite_width: -1,
      inpaint_strength: 0.5,
      overwrite_height: -1,
      overwrite_switch: -1,
      style_selections: "Fooocus V2,Fooocus Enhance,Fooocus Sharp",
      loras_custom_urls: "",
      uov_upscale_value: 0,
      use_default_loras: true,
      adm_scaler_negative: 0.8,
      adm_scaler_positive: 1.5,
      canny_low_threshold: 64,
      controlnet_softness: 0.25,
      outpaint_selections: "",
      canny_high_threshold: 128,
      invert_mask_checkbox: false,
      outpaint_distance_top: 0,
      performance_selection: "Speed",
      outpaint_distance_left: 0,
      aspect_ratios_selection: "1152*896",
      inpaint_erode_or_dilate: 0,
      outpaint_distance_right: 0,
      overwrite_vary_strength: -1,
      inpaint_respective_field: 1,
      outpaint_distance_bottom: 0,
      skipping_cn_preprocessor: false,
      debugging_cn_preprocessor: false,
      inpaint_additional_prompt: "",
      overwrite_upscale_strength: -1,
      debugging_inpaint_preprocessor: false,
      inpaint_disable_initial_latent: false,
      mixing_image_prompt_and_inpaint: false,
      mixing_image_prompt_and_vary_upscale: false
    }
  }
);

console.log(output);

To learn more, take a look at the guide on getting started with Node.js.

Install Replicate’s Python client library:

pip install replicate

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Import the client:

import replicate

Run mlscade/fooocus using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "mlscade/fooocus:d2d177eb70e517bc90576e973967d23868b40c69ad10b4c536f3de3d17db14c2",
    input={
        "prompt": "",
        "cn_type1": "ImagePrompt",
        "cn_type2": "ImagePrompt",
        "cn_type3": "ImagePrompt",
        "cn_type4": "ImagePrompt",
        "sharpness": 2,
        "image_seed": -1,
        "uov_method": "Disabled",
        "adaptive_cfg": 7,
        "image_number": 1,
        "sampler_name": "dpmpp_2m_sde_gpu",
        "adm_scaler_end": 0.3,
        "guidance_scale": 7,
        "inpaint_engine": "v2.6",
        "overwrite_step": -1,
        "refiner_switch": 0.5,
        "scheduler_name": "karras",
        "negative_prompt": "",
        "overwrite_width": -1,
        "inpaint_strength": 0.5,
        "overwrite_height": -1,
        "overwrite_switch": -1,
        "style_selections": "Fooocus V2,Fooocus Enhance,Fooocus Sharp",
        "loras_custom_urls": "",
        "uov_upscale_value": 0,
        "use_default_loras": True,
        "adm_scaler_negative": 0.8,
        "adm_scaler_positive": 1.5,
        "canny_low_threshold": 64,
        "controlnet_softness": 0.25,
        "outpaint_selections": "",
        "canny_high_threshold": 128,
        "invert_mask_checkbox": False,
        "outpaint_distance_top": 0,
        "performance_selection": "Speed",
        "outpaint_distance_left": 0,
        "aspect_ratios_selection": "1152*896",
        "inpaint_erode_or_dilate": 0,
        "outpaint_distance_right": 0,
        "overwrite_vary_strength": -1,
        "inpaint_respective_field": 1,
        "outpaint_distance_bottom": 0,
        "skipping_cn_preprocessor": False,
        "debugging_cn_preprocessor": False,
        "inpaint_additional_prompt": "",
        "overwrite_upscale_strength": -1,
        "debugging_inpaint_preprocessor": False,
        "inpaint_disable_initial_latent": False,
        "mixing_image_prompt_and_inpaint": False,
        "mixing_image_prompt_and_vary_upscale": False
    }
)
print(output)

To learn more, take a look at the guide on getting started with Python.

Set the REPLICATE_API_TOKEN environment variable:

export REPLICATE_API_TOKEN=<paste-your-token-here>

Find your API token in your account settings.

Run mlscade/fooocus using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

curl -s -X POST \
  -H "Authorization: Bearer $REPLICATE_API_TOKEN" \
  -H "Content-Type: application/json" \
  -H "Prefer: wait" \
  -d $'{
    "version": "mlscade/fooocus:d2d177eb70e517bc90576e973967d23868b40c69ad10b4c536f3de3d17db14c2",
    "input": {
      "prompt": "",
      "cn_type1": "ImagePrompt",
      "cn_type2": "ImagePrompt",
      "cn_type3": "ImagePrompt",
      "cn_type4": "ImagePrompt",
      "sharpness": 2,
      "image_seed": -1,
      "uov_method": "Disabled",
      "adaptive_cfg": 7,
      "image_number": 1,
      "sampler_name": "dpmpp_2m_sde_gpu",
      "adm_scaler_end": 0.3,
      "guidance_scale": 7,
      "inpaint_engine": "v2.6",
      "overwrite_step": -1,
      "refiner_switch": 0.5,
      "scheduler_name": "karras",
      "negative_prompt": "",
      "overwrite_width": -1,
      "inpaint_strength": 0.5,
      "overwrite_height": -1,
      "overwrite_switch": -1,
      "style_selections": "Fooocus V2,Fooocus Enhance,Fooocus Sharp",
      "loras_custom_urls": "",
      "uov_upscale_value": 0,
      "use_default_loras": true,
      "adm_scaler_negative": 0.8,
      "adm_scaler_positive": 1.5,
      "canny_low_threshold": 64,
      "controlnet_softness": 0.25,
      "outpaint_selections": "",
      "canny_high_threshold": 128,
      "invert_mask_checkbox": false,
      "outpaint_distance_top": 0,
      "performance_selection": "Speed",
      "outpaint_distance_left": 0,
      "aspect_ratios_selection": "1152*896",
      "inpaint_erode_or_dilate": 0,
      "outpaint_distance_right": 0,
      "overwrite_vary_strength": -1,
      "inpaint_respective_field": 1,
      "outpaint_distance_bottom": 0,
      "skipping_cn_preprocessor": false,
      "debugging_cn_preprocessor": false,
      "inpaint_additional_prompt": "",
      "overwrite_upscale_strength": -1,
      "debugging_inpaint_preprocessor": false,
      "inpaint_disable_initial_latent": false,
      "mixing_image_prompt_and_inpaint": false,
      "mixing_image_prompt_and_vary_upscale": false
    }
  }' \
  https://api.replicate.com/v1/predictions

To learn more, take a look at Replicate’s HTTP API reference docs.

Output

No output yet! Press "Submit" to start a prediction.

Run time and cost

This model costs approximately $0.17 to run on Replicate, or 5 runs per $1, but this varies depending on your inputs. It is also open source and you can run it on your own computer with Docker.

This model runs on Nvidia A100 (80GB) GPU hardware. Predictions typically complete within 124 seconds. The predict time for this model varies significantly based on the inputs.

Readme

This model doesn't have a readme.