fofr/realvisxl-v3-multi-controlnet-lora – Run with an API on Replicate

fofr / realvisxl-v3-multi-controlnet-lora

RealVisXl V3 with multi-controlnet, lora loading, img2img, inpainting

Warm

Public
1.3M runs
L40S
GitHub
Paper
License

Run with an API

Playground API Examples README Versions

Input

prompt

string

Shift + Return to add a new line

A detailed photo of an astronaut riding a unicorn through a field of flowersA detailed photo of an astronaut riding a unicorn through a field of flowers

Input prompt

Default: "An astronaut riding a rainbow unicorn"

negative_prompt

string

Shift + Return to add a new line

Negative Prompt

Default: ""

image

file

Input image for img2img or inpaint mode

mask

file

Input mask for inpaint mode. Black areas will be preserved, white areas will be inpainted.

width

integer

Width of output image

Default: 768

height

integer

Height of output image

Default: 768

sizing_strategy

string

Decide how to resize images – use width/height, resize based on input image or control image

Default: "width_height"

num_outputs

integer

(minimum: 1, maximum: 4)

Number of images to output

Default: 1

scheduler

string

scheduler

Default: "K_EULER"

num_inference_steps

integer

(minimum: 1, maximum: 500)

Number of denoising steps

Default: 30

guidance_scale

number

(minimum: 1, maximum: 50)

Scale for classifier-free guidance

Default: 7.5

prompt_strength

number

(minimum: 0, maximum: 1)

Prompt strength when using img2img / inpaint. 1.0 corresponds to full destruction of information in image

Default: 0.8

seed

integer

Random seed. Leave blank to randomize the seed

refine

string

Which refine style to use

Default: "no_refiner"

refine_steps

integer

For base_image_refiner, the number of steps to refine, defaults to num_inference_steps

apply_watermark

boolean

Applies a watermark to enable determining if an image is generated in downstream applications. If you have other provisions for generating or deploying images safely, you can use this to disable watermarking.

Default: false

lora_scale

number

(minimum: 0, maximum: 1)

LoRA additive scale. Only applicable on trained models.

Default: 0.6

lora_weights

string

Shift + Return to add a new line

Replicate LoRA weights to use. Leave blank to use the default weights.

disable_safety_checker

boolean

Disable safety checker for generated images. This feature is only available through the API.

Default: false

controlnet_1

string

Controlnet

Default: "none"

controlnet_1_image

file

Preview

Input image for first controlnet

controlnet_1_conditioning_scale

number

(minimum: 0, maximum: 4)

How strong the controlnet conditioning is

Default: 0.75

controlnet_1_start

number

(minimum: 0, maximum: 1)

When controlnet conditioning starts

Default: 0

controlnet_1_end

number

(minimum: 0, maximum: 1)

When controlnet conditioning ends

Default: 1

controlnet_2

string

Controlnet

Default: "none"

controlnet_2_image

file

Input image for second controlnet

controlnet_2_conditioning_scale

number

(minimum: 0, maximum: 4)

How strong the controlnet conditioning is

Default: 0.75

controlnet_2_start

number

(minimum: 0, maximum: 1)

When controlnet conditioning starts

Default: 0

controlnet_2_end

number

(minimum: 0, maximum: 1)

When controlnet conditioning ends

Default: 1

controlnet_3

string

Controlnet

Default: "none"

controlnet_3_image

file

Input image for third controlnet

controlnet_3_conditioning_scale

number

(minimum: 0, maximum: 4)

How strong the controlnet conditioning is

Default: 0.75

controlnet_3_start

number

(minimum: 0, maximum: 1)

When controlnet conditioning starts

Default: 0

controlnet_3_end

number

(minimum: 0, maximum: 1)

When controlnet conditioning ends

Default: 1

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

npx create-replicate --model=fofr/realvisxl-v3-multi-controlnet-lora

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 fofr/realvisxl-v3-multi-controlnet-lora using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "fofr/realvisxl-v3-multi-controlnet-lora:90a4a3604cd637cb9f1a2bdae1cfa9ed869362ca028814cdce310a78e27daade",
  {
    input: {
      width: 768,
      height: 768,
      prompt: "A detailed photo of an astronaut riding a unicorn through a field of flowers",
      refine: "no_refiner",
      scheduler: "K_EULER",
      lora_scale: 0.8,
      num_outputs: 1,
      controlnet_1: "soft_edge_hed",
      controlnet_2: "none",
      controlnet_3: "none",
      guidance_scale: 7.5,
      apply_watermark: false,
      negative_prompt: "",
      prompt_strength: 0.8,
      sizing_strategy: "width_height",
      controlnet_1_end: 1,
      controlnet_2_end: 1,
      controlnet_3_end: 1,
      controlnet_1_image: "https://replicate.delivery/pbxt/JsfCQE8k1lsCinW1yo76yKMQe6R5MRt9WLL3H5T5Ypc5wasq/020e656d-0c71-45c3-a7f5-1facf7d52d4f.png",
      controlnet_1_start: 0,
      controlnet_2_start: 0,
      controlnet_3_start: 0,
      num_inference_steps: 30,
      controlnet_1_conditioning_scale: 0.8,
      controlnet_2_conditioning_scale: 0.8,
      controlnet_3_conditioning_scale: 0.75
    }
  }
);
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 fofr/realvisxl-v3-multi-controlnet-lora using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "fofr/realvisxl-v3-multi-controlnet-lora:90a4a3604cd637cb9f1a2bdae1cfa9ed869362ca028814cdce310a78e27daade",
    input={
        "width": 768,
        "height": 768,
        "prompt": "A detailed photo of an astronaut riding a unicorn through a field of flowers",
        "refine": "no_refiner",
        "scheduler": "K_EULER",
        "lora_scale": 0.8,
        "num_outputs": 1,
        "controlnet_1": "soft_edge_hed",
        "controlnet_2": "none",
        "controlnet_3": "none",
        "guidance_scale": 7.5,
        "apply_watermark": False,
        "negative_prompt": "",
        "prompt_strength": 0.8,
        "sizing_strategy": "width_height",
        "controlnet_1_end": 1,
        "controlnet_2_end": 1,
        "controlnet_3_end": 1,
        "controlnet_1_image": "https://replicate.delivery/pbxt/JsfCQE8k1lsCinW1yo76yKMQe6R5MRt9WLL3H5T5Ypc5wasq/020e656d-0c71-45c3-a7f5-1facf7d52d4f.png",
        "controlnet_1_start": 0,
        "controlnet_2_start": 0,
        "controlnet_3_start": 0,
        "num_inference_steps": 30,
        "controlnet_1_conditioning_scale": 0.8,
        "controlnet_2_conditioning_scale": 0.8,
        "controlnet_3_conditioning_scale": 0.75
    }
)
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 fofr/realvisxl-v3-multi-controlnet-lora 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": "90a4a3604cd637cb9f1a2bdae1cfa9ed869362ca028814cdce310a78e27daade",
    "input": {
      "width": 768,
      "height": 768,
      "prompt": "A detailed photo of an astronaut riding a unicorn through a field of flowers",
      "refine": "no_refiner",
      "scheduler": "K_EULER",
      "lora_scale": 0.8,
      "num_outputs": 1,
      "controlnet_1": "soft_edge_hed",
      "controlnet_2": "none",
      "controlnet_3": "none",
      "guidance_scale": 7.5,
      "apply_watermark": false,
      "negative_prompt": "",
      "prompt_strength": 0.8,
      "sizing_strategy": "width_height",
      "controlnet_1_end": 1,
      "controlnet_2_end": 1,
      "controlnet_3_end": 1,
      "controlnet_1_image": "https://replicate.delivery/pbxt/JsfCQE8k1lsCinW1yo76yKMQe6R5MRt9WLL3H5T5Ypc5wasq/020e656d-0c71-45c3-a7f5-1facf7d52d4f.png",
      "controlnet_1_start": 0,
      "controlnet_2_start": 0,
      "controlnet_3_start": 0,
      "num_inference_steps": 30,
      "controlnet_1_conditioning_scale": 0.8,
      "controlnet_2_conditioning_scale": 0.8,
      "controlnet_3_conditioning_scale": 0.75
    }
  }' \
  https://api.replicate.com/v1/predictions

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

Output

{
  "completed_at": "2024-01-05T15:22:42.927298Z",
  "created_at": "2024-01-05T15:22:33.560461Z",
  "data_removed": false,
  "error": null,
  "id": "66toqj3bmu6zvqr4bj2hlmbsmu",
  "input": {
    "width": 768,
    "height": 768,
    "prompt": "A detailed photo of an astronaut riding a unicorn through a field of flowers",
    "refine": "no_refiner",
    "scheduler": "K_EULER",
    "lora_scale": 0.8,
    "num_outputs": 1,
    "controlnet_1": "soft_edge_hed",
    "controlnet_2": "none",
    "controlnet_3": "none",
    "guidance_scale": 7.5,
    "apply_watermark": false,
    "negative_prompt": "",
    "prompt_strength": 0.8,
    "sizing_strategy": "width_height",
    "controlnet_1_end": 1,
    "controlnet_2_end": 1,
    "controlnet_3_end": 1,
    "controlnet_1_image": "https://replicate.delivery/pbxt/JsfCQE8k1lsCinW1yo76yKMQe6R5MRt9WLL3H5T5Ypc5wasq/020e656d-0c71-45c3-a7f5-1facf7d52d4f.png",
    "controlnet_1_start": 0,
    "controlnet_2_start": 0,
    "controlnet_3_start": 0,
    "num_inference_steps": 30,
    "controlnet_1_conditioning_scale": 0.8,
    "controlnet_2_conditioning_scale": 0.8,
    "controlnet_3_conditioning_scale": 0.75
  },
  "logs": "Using seed: 19498\nUsing given dimensions\nresize took: 0.02s\nPrompt: A detailed photo of an astronaut riding a unicorn through a field of flowers\nProcessing image with soft_edge_hed\ncontrolnet preprocess took: 0.43s\nUsing txt2img + controlnet pipeline\nLoading pipeline components...:   0%|          | 0/7 [00:00<?, ?it/s]\nLoading pipeline components...: 100%|██████████| 7/7 [00:00<00:00, 15485.30it/s]\nYou have 1 ControlNets and you have passed 1 prompts. The conditionings will be fixed across the prompts.\n  0%|          | 0/30 [00:00<?, ?it/s]\n  3%|▎         | 1/30 [00:00<00:06,  4.63it/s]\n  7%|▋         | 2/30 [00:00<00:06,  4.62it/s]\n 10%|█         | 3/30 [00:00<00:05,  4.62it/s]\n 13%|█▎        | 4/30 [00:00<00:05,  4.62it/s]\n 17%|█▋        | 5/30 [00:01<00:05,  4.63it/s]\n 20%|██        | 6/30 [00:01<00:05,  4.63it/s]\n 23%|██▎       | 7/30 [00:01<00:04,  4.62it/s]\n 27%|██▋       | 8/30 [00:01<00:04,  4.62it/s]\n 30%|███       | 9/30 [00:01<00:04,  4.63it/s]\n 33%|███▎      | 10/30 [00:02<00:04,  4.63it/s]\n 37%|███▋      | 11/30 [00:02<00:04,  4.62it/s]\n 40%|████      | 12/30 [00:02<00:03,  4.62it/s]\n 43%|████▎     | 13/30 [00:02<00:03,  4.62it/s]\n 47%|████▋     | 14/30 [00:03<00:03,  4.62it/s]\n 50%|█████     | 15/30 [00:03<00:03,  4.62it/s]\n 53%|█████▎    | 16/30 [00:03<00:03,  4.62it/s]\n 57%|█████▋    | 17/30 [00:03<00:02,  4.62it/s]\n 60%|██████    | 18/30 [00:03<00:02,  4.62it/s]\n 63%|██████▎   | 19/30 [00:04<00:02,  4.62it/s]\n 67%|██████▋   | 20/30 [00:04<00:02,  4.62it/s]\n 70%|███████   | 21/30 [00:04<00:01,  4.62it/s]\n 73%|███████▎  | 22/30 [00:04<00:01,  4.61it/s]\n 77%|███████▋  | 23/30 [00:04<00:01,  4.62it/s]\n 80%|████████  | 24/30 [00:05<00:01,  4.62it/s]\n 83%|████████▎ | 25/30 [00:05<00:01,  4.62it/s]\n 87%|████████▋ | 26/30 [00:05<00:00,  4.62it/s]\n 90%|█████████ | 27/30 [00:05<00:00,  4.61it/s]\n 93%|█████████▎| 28/30 [00:06<00:00,  4.61it/s]\n 97%|█████████▋| 29/30 [00:06<00:00,  4.61it/s]\n100%|██████████| 30/30 [00:06<00:00,  4.61it/s]\n100%|██████████| 30/30 [00:06<00:00,  4.62it/s]\ninference took: 6.76s\nprediction took: 7.43s",
  "metrics": {
    "predict_time": 9.328726,
    "total_time": 9.366837
  },
  "output": [
    "https://replicate.delivery/pbxt/2hRD2xil2dq5IxJz0wC403uh1ckANw7srVhHxbPfn52ghuEJA/control-0.png",
    "https://replicate.delivery/pbxt/mUtp8mKk8yI0EJ5olzsnpkeTbAcmy2OTEqnXXc8EFGLhhuEJA/out-0.png"
  ],
  "started_at": "2024-01-05T15:22:33.598572Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/66toqj3bmu6zvqr4bj2hlmbsmu",
    "cancel": "https://api.replicate.com/v1/predictions/66toqj3bmu6zvqr4bj2hlmbsmu/cancel"
  },
  "version": "90a4a3604cd637cb9f1a2bdae1cfa9ed869362ca028814cdce310a78e27daade"
}

Generated in

9.3 seconds

Tweak it Report

Using seed: 19498
Using given dimensions
resize took: 0.02s
Prompt: A detailed photo of an astronaut riding a unicorn through a field of flowers
Processing image with soft_edge_hed
controlnet preprocess took: 0.43s
Using txt2img + controlnet pipeline
Loading pipeline components...:   0%|          | 0/7 [00:00<?, ?it/s]
Loading pipeline components...: 100%|██████████| 7/7 [00:00<00:00, 15485.30it/s]
You have 1 ControlNets and you have passed 1 prompts. The conditionings will be fixed across the prompts.
  0%|          | 0/30 [00:00<?, ?it/s]
  3%|▎         | 1/30 [00:00<00:06,  4.63it/s]
  7%|▋         | 2/30 [00:00<00:06,  4.62it/s]
 10%|█         | 3/30 [00:00<00:05,  4.62it/s]
 13%|█▎        | 4/30 [00:00<00:05,  4.62it/s]
 17%|█▋        | 5/30 [00:01<00:05,  4.63it/s]
 20%|██        | 6/30 [00:01<00:05,  4.63it/s]
 23%|██▎       | 7/30 [00:01<00:04,  4.62it/s]
 27%|██▋       | 8/30 [00:01<00:04,  4.62it/s]
 30%|███       | 9/30 [00:01<00:04,  4.63it/s]
 33%|███▎      | 10/30 [00:02<00:04,  4.63it/s]
 37%|███▋      | 11/30 [00:02<00:04,  4.62it/s]
 40%|████      | 12/30 [00:02<00:03,  4.62it/s]
 43%|████▎     | 13/30 [00:02<00:03,  4.62it/s]
 47%|████▋     | 14/30 [00:03<00:03,  4.62it/s]
 50%|█████     | 15/30 [00:03<00:03,  4.62it/s]
 53%|█████▎    | 16/30 [00:03<00:03,  4.62it/s]
 57%|█████▋    | 17/30 [00:03<00:02,  4.62it/s]
 60%|██████    | 18/30 [00:03<00:02,  4.62it/s]
 63%|██████▎   | 19/30 [00:04<00:02,  4.62it/s]
 67%|██████▋   | 20/30 [00:04<00:02,  4.62it/s]
 70%|███████   | 21/30 [00:04<00:01,  4.62it/s]
 73%|███████▎  | 22/30 [00:04<00:01,  4.61it/s]
 77%|███████▋  | 23/30 [00:04<00:01,  4.62it/s]
 80%|████████  | 24/30 [00:05<00:01,  4.62it/s]
 83%|████████▎ | 25/30 [00:05<00:01,  4.62it/s]
 87%|████████▋ | 26/30 [00:05<00:00,  4.62it/s]
 90%|█████████ | 27/30 [00:05<00:00,  4.61it/s]
 93%|█████████▎| 28/30 [00:06<00:00,  4.61it/s]
 97%|█████████▋| 29/30 [00:06<00:00,  4.61it/s]
100%|██████████| 30/30 [00:06<00:00,  4.61it/s]
100%|██████████| 30/30 [00:06<00:00,  4.62it/s]
inference took: 6.76s
prediction took: 7.43s

Examples

View more examples

Run time and cost

This model costs approximately $0.017 to run on Replicate, or 58 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 L40S GPU hardware. Predictions typically complete within 18 seconds.

Readme

An SDXL full model tune aimed at photorealism. https://huggingface.co/SG161222/RealVisXL_V3.0

RealVis XL V3.0 with:

img2img
inpainting
custom Replicate lora loading
up to 3 simultaneous controlnets with different images
img2img plus controlnet
inpainting plus controlnet
controlnet conditioning strengths
controlnet start and end controls
SDXL refiner
Image resizing based on width/height, input image or a control image
Disable safety checker via API

Controlnets included:

canny
midas depth
leres depth
soft edge hed
soft edge pidi
openpose
QR Monster (illusions)
lineart
lineart anime