okaris/omni-bg – Run with an API on Replicate

okaris / omni-bg

Cold

Public
49 runs
A100 (80GB)

Run with an API

Playground API Examples README Versions

Input

job_id

string

Shift + Return to add a new line

Job ID for the model. Randomly generated if not provided

image

*file

Base image for the model

subject

*string

Shift + Return to add a new line

Subject of the image

placed_on

*string

Shift + Return to add a new line

Where the subject is placed

background

*string

Shift + Return to add a new line

Background description

camera_orientation

*string

Shift + Return to add a new line

Camera orientation

camera_angle

*string

Shift + Return to add a new line

Camera angle

main_light_direction

*string

Shift + Return to add a new line

Main light direction

proposed_background

*string

Shift + Return to add a new line

Proposed background

proposed_brightness

*number

Proposed brightness

Including prompt_suffix and 30 more...

prompt_suffix

string

Shift + Return to add a new line

,high quality, medium format, f/32, realistic, cinematic, RAW,high quality, medium format, f/32, realistic, cinematic, RAW

Prompt suffix

Default: ",high quality, medium format, f/32, realistic, cinematic, RAW"

prompt_prefix

string

Shift + Return to add a new line

Starting prompt for the model. Gets added with other variables as needed

Default: "photograph"

negative_prompt

string

Shift + Return to add a new line

Negative prompt for the model. Not necessary, but can be used to guide the model

Default: ""

override_prompt

string

Shift + Return to add a new line

Override prompt for the model. Not necessary, but can be used to guide the model

Default: ""

background_brightness

number

(minimum: -1, maximum: 1)

Background brightness for the model

Default: 0.5

force_center

boolean

Force the subject to be centered

Default: false

seed

integer

Random seed for the model. Leave as -1 for random seed

Default: -1

height

integer

Height of the output image. Must be divisible by 32, but this is handled automatically

Default: 768

width

integer

Width of the output image. Must be divisible by 32, but this is handled automatically

Default: 1024

final_step_upscale_4x

boolean

Final step upscale 4x for the model

Default: false

debug_mode

boolean

Debug mode for the model

Default: false

object_padding

number

(minimum: 0, maximum: 1)

Object padding for the model. Percent of the object width

Default: 0.1

first_pass_steps

integer

(minimum: 10, maximum: 50)

Number of steps for the first pass

Default: 10

first_pass_i2i_strength

number

(minimum: 0, maximum: 1)

I2I strength for the first pass. Must be 1.0 unless you feel like experimenting

Default: 1

first_pass_pag_scale

number

(minimum: 0.5, maximum: 5)

PAG scale for the first pass. 1 or 2 seem to work best

Default: 2

first_pass_guidance_scale

number

(minimum: 1, maximum: 14)

Guidance scale for the first pass

Default: 4

first_pass_outpaint_strength

number

(minimum: 0, maximum: 1)

Outpaint strength for the first pass

Default: 1

first_pass_depth_strength

number

(minimum: 0, maximum: 1)

Depth strength for the first pass. For the first pass we want more flexibility, so 0.0 is a good starting point

Default: 0

first_pass_line_strength

number

(minimum: 0, maximum: 1)

Line strength for the first pass

Default: 1

skip_second_pass

boolean

Skip the second pass

Default: false

second_pass_steps

integer

(minimum: 10, maximum: 50)

Number of steps for the second pass

Default: 10

second_pass_i2i_strength

number

(minimum: 0, maximum: 1)

I2I strength for the second pass

Default: 0.5

second_pass_pag_scale

number

(minimum: 0.5, maximum: 5)

PAG scale for the second pass

Default: 1

second_pass_guidance_scale

number

(minimum: 1, maximum: 14)

Guidance scale for the second pass

Default: 4

second_pass_outpaint_strength

number

(minimum: 0, maximum: 1)

Outpaint strength for the second pass

Default: 1

second_pass_depth_strength

number

(minimum: 0, maximum: 1)

Depth strength for the second pass. We now want to lock in the depth, so 1.0 is a good starting point

Default: 1

second_pass_line_strength

number

(minimum: 0, maximum: 1)

Line strength for the second pass

Default: 1

skip_light_pass

boolean

Skip the light pass

Default: false

light_pass_steps

integer

(minimum: 10, maximum: 50)

Number of steps for the light pass. This is a two step process and takes 1.9x steps

Default: 10

light_pass_guidance_scale

number

(minimum: 1, maximum: 14)

Guidance scale for the light pass

Default: 7

light_mix_scale

number

(minimum: 0, maximum: 1)

Mix scale for the light pass. 0.5 is a good starting point

Default: 0.5

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

npx create-replicate --model=okaris/omni-bg

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 okaris/omni-bg using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "okaris/omni-bg:7ac428cb675977ea6b2eac6478174b4c4554692e5d36bf5e01f0e61b7a3ec0b7",
  {
    input: {
      seed: -1,
      width: 1024,
      height: 768,
      debug_mode: false,
      force_center: false,
      prompt_prefix: "photograph",
      prompt_suffix: ",high quality, medium format, f/32, realistic, cinematic, RAW",
      object_padding: 0.1,
      light_mix_scale: 0.5,
      negative_prompt: "",
      override_prompt: "",
      skip_light_pass: false,
      first_pass_steps: 10,
      light_pass_steps: 10,
      skip_second_pass: false,
      second_pass_steps: 10,
      first_pass_pag_scale: 2,
      background_brightness: 0.5,
      final_step_upscale_4x: false,
      second_pass_pag_scale: 1,
      first_pass_i2i_strength: 1,
      first_pass_line_strength: 1,
      second_pass_i2i_strength: 0.5,
      first_pass_depth_strength: 0,
      first_pass_guidance_scale: 4,
      light_pass_guidance_scale: 7,
      second_pass_line_strength: 1,
      second_pass_depth_strength: 1,
      second_pass_guidance_scale: 4,
      first_pass_outpaint_strength: 1,
      second_pass_outpaint_strength: 1
    }
  }
);

// To access the file URL:
console.log(output[0].url()); //=> "http://example.com"

// To write the file to disk:
fs.writeFile("my-image.png", output[0]);

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 okaris/omni-bg using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "okaris/omni-bg:7ac428cb675977ea6b2eac6478174b4c4554692e5d36bf5e01f0e61b7a3ec0b7",
    input={
        "seed": -1,
        "width": 1024,
        "height": 768,
        "debug_mode": False,
        "force_center": False,
        "prompt_prefix": "photograph",
        "prompt_suffix": ",high quality, medium format, f/32, realistic, cinematic, RAW",
        "object_padding": 0.1,
        "light_mix_scale": 0.5,
        "negative_prompt": "",
        "override_prompt": "",
        "skip_light_pass": False,
        "first_pass_steps": 10,
        "light_pass_steps": 10,
        "skip_second_pass": False,
        "second_pass_steps": 10,
        "first_pass_pag_scale": 2,
        "background_brightness": 0.5,
        "final_step_upscale_4x": False,
        "second_pass_pag_scale": 1,
        "first_pass_i2i_strength": 1,
        "first_pass_line_strength": 1,
        "second_pass_i2i_strength": 0.5,
        "first_pass_depth_strength": 0,
        "first_pass_guidance_scale": 4,
        "light_pass_guidance_scale": 7,
        "second_pass_line_strength": 1,
        "second_pass_depth_strength": 1,
        "second_pass_guidance_scale": 4,
        "first_pass_outpaint_strength": 1,
        "second_pass_outpaint_strength": 1
    }
)

# The okaris/omni-bg model can stream output as it's running.
# The predict method returns an iterator, and you can iterate over that output.
for item in output:
    # https://replicate.com/okaris/omni-bg/api#output-schema
    print(item)

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 okaris/omni-bg 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": "7ac428cb675977ea6b2eac6478174b4c4554692e5d36bf5e01f0e61b7a3ec0b7",
    "input": {
      "seed": -1,
      "width": 1024,
      "height": 768,
      "debug_mode": false,
      "force_center": false,
      "prompt_prefix": "photograph",
      "prompt_suffix": ",high quality, medium format, f/32, realistic, cinematic, RAW",
      "object_padding": 0.1,
      "light_mix_scale": 0.5,
      "negative_prompt": "",
      "override_prompt": "",
      "skip_light_pass": false,
      "first_pass_steps": 10,
      "light_pass_steps": 10,
      "skip_second_pass": false,
      "second_pass_steps": 10,
      "first_pass_pag_scale": 2,
      "background_brightness": 0.5,
      "final_step_upscale_4x": false,
      "second_pass_pag_scale": 1,
      "first_pass_i2i_strength": 1,
      "first_pass_line_strength": 1,
      "second_pass_i2i_strength": 0.5,
      "first_pass_depth_strength": 0,
      "first_pass_guidance_scale": 4,
      "light_pass_guidance_scale": 7,
      "second_pass_line_strength": 1,
      "second_pass_depth_strength": 1,
      "second_pass_guidance_scale": 4,
      "first_pass_outpaint_strength": 1,
      "second_pass_outpaint_strength": 1
    }
  }' \
  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 runs on Nvidia A100 (80GB) GPU hardware. We don't yet have enough runs of this model to provide performance information.

Readme

This model doesn't have a readme.