fermatresearch/sdxl-controlnet-lora – Run with an API on Replicate

fermatresearch / sdxl-controlnet-lora

'''Last update: Now supports img2img.''' SDXL Canny controlnet with LoRA support.

Cold

Public
896K runs
L40S
GitHub

Iterate in playground

Run with an API

Playground API Examples README Versions

Input

prompt

string

Shift + Return to add a new line

shot in the style of sksfer, a woman in alaskashot in the style of sksfer, a woman in alaska

Input prompt

Default: "An astronaut riding a rainbow unicorn"

image

file

Input image for img2img or inpaint mode

img2img

boolean

Use img2img pipeline, it will use the image input both as the control image and the base image.

condition_scale

number

(minimum: 0, maximum: 2)

The bigger this number is, the more ControlNet interferes

Default: 1.1

strength

number

(minimum: 0, maximum: 1)

When img2img is active, the denoising strength. 1 means total destruction of the input image.

Default: 0.8

negative_prompt

string

Shift + Return to add a new line

Input Negative Prompt

Default: ""

num_inference_steps

integer

(minimum: 1, maximum: 500)

Number of denoising steps

Default: 30

num_outputs

integer

(minimum: 1, maximum: 4)

Number of images to output

Default: 1

scheduler

string

scheduler

Default: "K_EULER"

guidance_scale

number

(minimum: 1, maximum: 50)

Scale for classifier-free guidance

Default: 7.5

seed

integer

Random seed. Leave blank to randomize the seed

refine

string

Whether to use refinement steps or not

Default: "base_image_refiner"

refine_steps

integer

For base_image_refiner, the number of steps to refine

Default: 10

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: true

lora_scale

number

(minimum: 0, maximum: 1)

LoRA additive scale. Only applicable on trained models.

Default: 0.95

lora_weights

string

Shift + Return to add a new line

https://pbxt.replicate.delivery/mwN3AFyYZyouOB03Uhw8ubKW9rpqMgdtL9zYV9GF2WGDiwbE/trained_model.tarhttps://pbxt.replicate.delivery/mwN3AFyYZyouOB03Uhw8ubKW9rpqMgdtL9zYV9GF2WGDiwbE/trained_model.tar

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

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

npx create-replicate --model=fermatresearch/sdxl-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";
import fs from "node:fs";

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

Run fermatresearch/sdxl-controlnet-lora using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "fermatresearch/sdxl-controlnet-lora:3bb13fe1c33c35987b33792b01b71ed6529d03f165d1c2416375859f09ca9fef",
  {
    input: {
      seed: null,
      image: "https://replicate.delivery/pbxt/JiOTMCHj4oGrTTf8Pg2r7vyI8YdXc5jL2IDyC2SfhuggjYe6/out-0%20%281%29.png",
      prompt: "shot in the style of sksfer, a woman in alaska",
      refine: "base_image_refiner",
      strength: 0.8,
      scheduler: "K_EULER",
      lora_scale: 0.95,
      num_outputs: 1,
      lora_weights: "https://pbxt.replicate.delivery/mwN3AFyYZyouOB03Uhw8ubKW9rpqMgdtL9zYV9GF2WGDiwbE/trained_model.tar",
      refine_steps: 20,
      guidance_scale: 7.5,
      apply_watermark: true,
      condition_scale: 0.5,
      negative_prompt: "",
      num_inference_steps: 40
    }
  }
);

// 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 fermatresearch/sdxl-controlnet-lora using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "fermatresearch/sdxl-controlnet-lora:3bb13fe1c33c35987b33792b01b71ed6529d03f165d1c2416375859f09ca9fef",
    input={
        "seed": null,
        "image": "https://replicate.delivery/pbxt/JiOTMCHj4oGrTTf8Pg2r7vyI8YdXc5jL2IDyC2SfhuggjYe6/out-0%20%281%29.png",
        "prompt": "shot in the style of sksfer, a woman in alaska",
        "refine": "base_image_refiner",
        "strength": 0.8,
        "scheduler": "K_EULER",
        "lora_scale": 0.95,
        "num_outputs": 1,
        "lora_weights": "https://pbxt.replicate.delivery/mwN3AFyYZyouOB03Uhw8ubKW9rpqMgdtL9zYV9GF2WGDiwbE/trained_model.tar",
        "refine_steps": 20,
        "guidance_scale": 7.5,
        "apply_watermark": True,
        "condition_scale": 0.5,
        "negative_prompt": "",
        "num_inference_steps": 40
    }
)
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 fermatresearch/sdxl-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": "fermatresearch/sdxl-controlnet-lora:3bb13fe1c33c35987b33792b01b71ed6529d03f165d1c2416375859f09ca9fef",
    "input": {
      "seed": null,
      "image": "https://replicate.delivery/pbxt/JiOTMCHj4oGrTTf8Pg2r7vyI8YdXc5jL2IDyC2SfhuggjYe6/out-0%20%281%29.png",
      "prompt": "shot in the style of sksfer, a woman in alaska",
      "refine": "base_image_refiner",
      "strength": 0.8,
      "scheduler": "K_EULER",
      "lora_scale": 0.95,
      "num_outputs": 1,
      "lora_weights": "https://pbxt.replicate.delivery/mwN3AFyYZyouOB03Uhw8ubKW9rpqMgdtL9zYV9GF2WGDiwbE/trained_model.tar",
      "refine_steps": 20,
      "guidance_scale": 7.5,
      "apply_watermark": true,
      "condition_scale": 0.5,
      "negative_prompt": "",
      "num_inference_steps": 40
    }
  }' \
  https://api.replicate.com/v1/predictions

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

You can run this model locally using Cog. First, install Cog:

brew install cog

If you don’t have Homebrew, there are other installation options available.

Run this to download the model and run it in your local environment:

cog predict r8.im/fermatresearch/sdxl-controlnet-lora@sha256:3bb13fe1c33c35987b33792b01b71ed6529d03f165d1c2416375859f09ca9fef \
  -i 'seed=null' \
  -i 'image="https://replicate.delivery/pbxt/JiOTMCHj4oGrTTf8Pg2r7vyI8YdXc5jL2IDyC2SfhuggjYe6/out-0%20%281%29.png"' \
  -i 'prompt="shot in the style of sksfer, a woman in alaska"' \
  -i 'refine="base_image_refiner"' \
  -i 'strength=0.8' \
  -i 'scheduler="K_EULER"' \
  -i 'lora_scale=0.95' \
  -i 'num_outputs=1' \
  -i 'lora_weights="https://pbxt.replicate.delivery/mwN3AFyYZyouOB03Uhw8ubKW9rpqMgdtL9zYV9GF2WGDiwbE/trained_model.tar"' \
  -i 'refine_steps=20' \
  -i 'guidance_scale=7.5' \
  -i 'apply_watermark=true' \
  -i 'condition_scale=0.5' \
  -i 'negative_prompt=""' \
  -i 'num_inference_steps=40'

To learn more, take a look at the Cog documentation.

Run this to download the model and run it in your local environment:

docker run -d -p 5000:5000 --gpus=all r8.im/fermatresearch/sdxl-controlnet-lora@sha256:3bb13fe1c33c35987b33792b01b71ed6529d03f165d1c2416375859f09ca9fef
curl -s -X POST \
  -H "Content-Type: application/json" \
  -d $'{
    "input": {
      "seed": null,
      "image": "https://replicate.delivery/pbxt/JiOTMCHj4oGrTTf8Pg2r7vyI8YdXc5jL2IDyC2SfhuggjYe6/out-0%20%281%29.png",
      "prompt": "shot in the style of sksfer, a woman in alaska",
      "refine": "base_image_refiner",
      "strength": 0.8,
      "scheduler": "K_EULER",
      "lora_scale": 0.95,
      "num_outputs": 1,
      "lora_weights": "https://pbxt.replicate.delivery/mwN3AFyYZyouOB03Uhw8ubKW9rpqMgdtL9zYV9GF2WGDiwbE/trained_model.tar",
      "refine_steps": 20,
      "guidance_scale": 7.5,
      "apply_watermark": true,
      "condition_scale": 0.5,
      "negative_prompt": "",
      "num_inference_steps": 40
    }
  }' \
  http://localhost:5000/predictions

To learn more, take a look at the Cog documentation.

Output

{
  "completed_at": "2023-10-17T13:48:50.759475Z",
  "created_at": "2023-10-17T13:48:30.378425Z",
  "data_removed": false,
  "error": null,
  "id": "ruzsnmtb7idrwecy2ue5ig6gli",
  "input": {
    "seed": null,
    "image": "https://replicate.delivery/pbxt/JiOTMCHj4oGrTTf8Pg2r7vyI8YdXc5jL2IDyC2SfhuggjYe6/out-0%20%281%29.png",
    "prompt": "shot in the style of sksfer, a woman in alaska",
    "refine": "base_image_refiner",
    "scheduler": "K_EULER",
    "lora_scale": 0.95,
    "num_outputs": 1,
    "lora_weights": "https://pbxt.replicate.delivery/mwN3AFyYZyouOB03Uhw8ubKW9rpqMgdtL9zYV9GF2WGDiwbE/trained_model.tar",
    "refine_steps": 20,
    "guidance_scale": 7.5,
    "apply_watermark": true,
    "condition_scale": 0.5,
    "negative_prompt": "",
    "num_inference_steps": 40
  },
  "logs": "Using seed: 44186\nloading custom weights\nweights already in cache\nLoading fine-tuned model\nDoes not have Unet. assume we are using LoRA\nLoading Unet LoRA\nPrompt: shot in the style of <s0><s1>, a woman in alaska\nOriginal width:1024, height:1024\nAspect Ratio: 1.00\nnew_width:1024, new_height:1024\ntxt2img mode\n  0%|          | 0/40 [00:00<?, ?it/s]\n  2%|▎         | 1/40 [00:00<00:14,  2.72it/s]\n  5%|▌         | 2/40 [00:00<00:13,  2.73it/s]\n  8%|▊         | 3/40 [00:01<00:13,  2.73it/s]\n 10%|█         | 4/40 [00:01<00:13,  2.73it/s]\n 12%|█▎        | 5/40 [00:01<00:12,  2.73it/s]\n 15%|█▌        | 6/40 [00:02<00:12,  2.73it/s]\n 18%|█▊        | 7/40 [00:02<00:12,  2.73it/s]\n 20%|██        | 8/40 [00:02<00:11,  2.73it/s]\n 22%|██▎       | 9/40 [00:03<00:11,  2.73it/s]\n 25%|██▌       | 10/40 [00:03<00:11,  2.72it/s]\n 28%|██▊       | 11/40 [00:04<00:10,  2.72it/s]\n 30%|███       | 12/40 [00:04<00:10,  2.72it/s]\n 32%|███▎      | 13/40 [00:04<00:09,  2.72it/s]\n 35%|███▌      | 14/40 [00:05<00:09,  2.72it/s]\n 38%|███▊      | 15/40 [00:05<00:09,  2.72it/s]\n 40%|████      | 16/40 [00:05<00:08,  2.72it/s]\n 42%|████▎     | 17/40 [00:06<00:08,  2.71it/s]\n 45%|████▌     | 18/40 [00:06<00:08,  2.71it/s]\n 48%|████▊     | 19/40 [00:06<00:07,  2.71it/s]\n 50%|█████     | 20/40 [00:07<00:07,  2.71it/s]\n 52%|█████▎    | 21/40 [00:07<00:06,  2.71it/s]\n 55%|█████▌    | 22/40 [00:08<00:06,  2.72it/s]\n 57%|█████▊    | 23/40 [00:08<00:06,  2.71it/s]\n 60%|██████    | 24/40 [00:08<00:05,  2.71it/s]\n 62%|██████▎   | 25/40 [00:09<00:05,  2.71it/s]\n 65%|██████▌   | 26/40 [00:09<00:05,  2.71it/s]\n 68%|██████▊   | 27/40 [00:09<00:04,  2.71it/s]\n 70%|███████   | 28/40 [00:10<00:04,  2.72it/s]\n 72%|███████▎  | 29/40 [00:10<00:04,  2.71it/s]\n 75%|███████▌  | 30/40 [00:11<00:03,  2.71it/s]\n 78%|███████▊  | 31/40 [00:11<00:03,  2.71it/s]\n 80%|████████  | 32/40 [00:11<00:02,  2.71it/s]\n 82%|████████▎ | 33/40 [00:12<00:02,  2.71it/s]\n 85%|████████▌ | 34/40 [00:12<00:02,  2.71it/s]\n 88%|████████▊ | 35/40 [00:12<00:01,  2.71it/s]\n 90%|█████████ | 36/40 [00:13<00:01,  2.71it/s]\n 92%|█████████▎| 37/40 [00:13<00:01,  2.71it/s]\n 95%|█████████▌| 38/40 [00:13<00:00,  2.71it/s]\n 98%|█████████▊| 39/40 [00:14<00:00,  2.71it/s]\n100%|██████████| 40/40 [00:14<00:00,  2.71it/s]\n100%|██████████| 40/40 [00:14<00:00,  2.72it/s]\n  0%|          | 0/6 [00:00<?, ?it/s]\n 17%|█▋        | 1/6 [00:00<00:01,  4.32it/s]\n 33%|███▎      | 2/6 [00:00<00:00,  4.30it/s]\n 50%|█████     | 3/6 [00:00<00:00,  4.30it/s]\n 67%|██████▋   | 4/6 [00:00<00:00,  4.28it/s]\n 83%|████████▎ | 5/6 [00:01<00:00,  4.28it/s]\n100%|██████████| 6/6 [00:01<00:00,  4.28it/s]\n100%|██████████| 6/6 [00:01<00:00,  4.29it/s]",
  "metrics": {
    "predict_time": 20.411215,
    "total_time": 20.38105
  },
  "output": [
    "https://replicate.delivery/pbxt/oDtYIK2lDoaKMtdE4E5ozQSa61BU3gc4aRvGF3xmFpdwCxbE/out-0.png"
  ],
  "started_at": "2023-10-17T13:48:30.348260Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/ruzsnmtb7idrwecy2ue5ig6gli",
    "cancel": "https://api.replicate.com/v1/predictions/ruzsnmtb7idrwecy2ue5ig6gli/cancel"
  },
  "version": "a4fb84022361602a2401d74435229e90da63ea4a2aab40ebf79afd7af5a081d4"
}

Generated in

20.4 seconds

Tweak itReport

Using seed: 44186
loading custom weights
weights already in cache
Loading fine-tuned model
Does not have Unet. assume we are using LoRA
Loading Unet LoRA
Prompt: shot in the style of <s0><s1>, a woman in alaska
Original width:1024, height:1024
Aspect Ratio: 1.00
new_width:1024, new_height:1024
txt2img mode
  0%|          | 0/40 [00:00<?, ?it/s]
  2%|▎         | 1/40 [00:00<00:14,  2.72it/s]
  5%|▌         | 2/40 [00:00<00:13,  2.73it/s]
  8%|▊         | 3/40 [00:01<00:13,  2.73it/s]
 10%|█         | 4/40 [00:01<00:13,  2.73it/s]
 12%|█▎        | 5/40 [00:01<00:12,  2.73it/s]
 15%|█▌        | 6/40 [00:02<00:12,  2.73it/s]
 18%|█▊        | 7/40 [00:02<00:12,  2.73it/s]
 20%|██        | 8/40 [00:02<00:11,  2.73it/s]
 22%|██▎       | 9/40 [00:03<00:11,  2.73it/s]
 25%|██▌       | 10/40 [00:03<00:11,  2.72it/s]
 28%|██▊       | 11/40 [00:04<00:10,  2.72it/s]
 30%|███       | 12/40 [00:04<00:10,  2.72it/s]
 32%|███▎      | 13/40 [00:04<00:09,  2.72it/s]
 35%|███▌      | 14/40 [00:05<00:09,  2.72it/s]
 38%|███▊      | 15/40 [00:05<00:09,  2.72it/s]
 40%|████      | 16/40 [00:05<00:08,  2.72it/s]
 42%|████▎     | 17/40 [00:06<00:08,  2.71it/s]
 45%|████▌     | 18/40 [00:06<00:08,  2.71it/s]
 48%|████▊     | 19/40 [00:06<00:07,  2.71it/s]
 50%|█████     | 20/40 [00:07<00:07,  2.71it/s]
 52%|█████▎    | 21/40 [00:07<00:06,  2.71it/s]
 55%|█████▌    | 22/40 [00:08<00:06,  2.72it/s]
 57%|█████▊    | 23/40 [00:08<00:06,  2.71it/s]
 60%|██████    | 24/40 [00:08<00:05,  2.71it/s]
 62%|██████▎   | 25/40 [00:09<00:05,  2.71it/s]
 65%|██████▌   | 26/40 [00:09<00:05,  2.71it/s]
 68%|██████▊   | 27/40 [00:09<00:04,  2.71it/s]
 70%|███████   | 28/40 [00:10<00:04,  2.72it/s]
 72%|███████▎  | 29/40 [00:10<00:04,  2.71it/s]
 75%|███████▌  | 30/40 [00:11<00:03,  2.71it/s]
 78%|███████▊  | 31/40 [00:11<00:03,  2.71it/s]
 80%|████████  | 32/40 [00:11<00:02,  2.71it/s]
 82%|████████▎ | 33/40 [00:12<00:02,  2.71it/s]
 85%|████████▌ | 34/40 [00:12<00:02,  2.71it/s]
 88%|████████▊ | 35/40 [00:12<00:01,  2.71it/s]
 90%|█████████ | 36/40 [00:13<00:01,  2.71it/s]
 92%|█████████▎| 37/40 [00:13<00:01,  2.71it/s]
 95%|█████████▌| 38/40 [00:13<00:00,  2.71it/s]
 98%|█████████▊| 39/40 [00:14<00:00,  2.71it/s]
100%|██████████| 40/40 [00:14<00:00,  2.71it/s]
100%|██████████| 40/40 [00:14<00:00,  2.72it/s]
  0%|          | 0/6 [00:00<?, ?it/s]
 17%|█▋        | 1/6 [00:00<00:01,  4.32it/s]
 33%|███▎      | 2/6 [00:00<00:00,  4.30it/s]
 50%|█████     | 3/6 [00:00<00:00,  4.30it/s]
 67%|██████▋   | 4/6 [00:00<00:00,  4.28it/s]
 83%|████████▎ | 5/6 [00:01<00:00,  4.28it/s]
100%|██████████| 6/6 [00:01<00:00,  4.28it/s]
100%|██████████| 6/6 [00:01<00:00,  4.29it/s]

This output was created using a different version of the model, fermatresearch/sdxl-controlnet-lora:a4fb8402.

Examples

View more examples

Run time and cost

This model costs approximately $0.011 to run on Replicate, or 90 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 12 seconds. The predict time for this model varies significantly based on the inputs.

Readme

This model doesn't have a readme.