fofr/style-transfer | Run with an API on Replicate

fofr / style-transfer

Transfer the style of one image to another

Warm

Public
965.9K runs
L40S
GitHub
License

Iterate in playground

Run with an API

Playground API Examples README Versions

Input

style_image

*file

Preview

Copy the style from this image

structure_image

file

An optional image to copy structure from. Output images will use the same aspect ratio.

prompt

string

Shift + Return to add a new line

Prompt for the image

Default: "An astronaut riding a unicorn"

negative_prompt

string

Shift + Return to add a new line

Things you do not want to see in your image

Default: ""

width

integer

Width of the output image (ignored if structure image given)

Default: 1024

height

integer

Height of the output image (ignored if structure image given)

Default: 1024

model

string

Model to use for the generation

Default: "fast"

number_of_images

integer

(minimum: 1, maximum: 10)

Number of images to generate

Default: 1

structure_depth_strength

number

(minimum: 0, maximum: 2)

Strength of the depth controlnet

Default: 1

structure_denoising_strength

number

(minimum: 0, maximum: 1)

How much of the original image (and colors) to preserve (0 is all, 1 is none, 0.65 is a good balance)

Default: 0.65

output_format

string

Format of the output images

Default: "webp"

output_quality

integer

(minimum: 0, maximum: 100)

Quality of the output images, from 0 to 100. 100 is best quality, 0 is lowest quality.

Default: 80

seed

integer

Set a seed for reproducibility. Random by default.

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

npx create-replicate --model=fofr/style-transfer

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 fofr/style-transfer using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "fofr/style-transfer:f1023890703bc0a5a3a2c21b5e498833be5f6ef6e70e9daf6b9b3a4fd8309cf0",
  {
    input: {
      model: "fast",
      width: 1024,
      height: 1024,
      prompt: "An astronaut riding a unicorn",
      style_image: "https://replicate.delivery/pbxt/KlTqluRakBzt7N5mm1WExEQCc4J3usa7E3n5dhttcayTqFRm/van-gogh.jpeg",
      output_format: "webp",
      output_quality: 80,
      negative_prompt: "",
      number_of_images: 1,
      structure_depth_strength: 1,
      structure_denoising_strength: 0.65
    }
  }
);

// 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 fofr/style-transfer using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "fofr/style-transfer:f1023890703bc0a5a3a2c21b5e498833be5f6ef6e70e9daf6b9b3a4fd8309cf0",
    input={
        "model": "fast",
        "width": 1024,
        "height": 1024,
        "prompt": "An astronaut riding a unicorn",
        "style_image": "https://replicate.delivery/pbxt/KlTqluRakBzt7N5mm1WExEQCc4J3usa7E3n5dhttcayTqFRm/van-gogh.jpeg",
        "output_format": "webp",
        "output_quality": 80,
        "negative_prompt": "",
        "number_of_images": 1,
        "structure_depth_strength": 1,
        "structure_denoising_strength": 0.65
    }
)
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/style-transfer 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": "fofr/style-transfer:f1023890703bc0a5a3a2c21b5e498833be5f6ef6e70e9daf6b9b3a4fd8309cf0",
    "input": {
      "model": "fast",
      "width": 1024,
      "height": 1024,
      "prompt": "An astronaut riding a unicorn",
      "style_image": "https://replicate.delivery/pbxt/KlTqluRakBzt7N5mm1WExEQCc4J3usa7E3n5dhttcayTqFRm/van-gogh.jpeg",
      "output_format": "webp",
      "output_quality": 80,
      "negative_prompt": "",
      "number_of_images": 1,
      "structure_depth_strength": 1,
      "structure_denoising_strength": 0.65
    }
  }' \
  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/fofr/style-transfer@sha256:f1023890703bc0a5a3a2c21b5e498833be5f6ef6e70e9daf6b9b3a4fd8309cf0 \
  -i 'model="fast"' \
  -i 'width=1024' \
  -i 'height=1024' \
  -i 'prompt="An astronaut riding a unicorn"' \
  -i 'style_image="https://replicate.delivery/pbxt/KlTqluRakBzt7N5mm1WExEQCc4J3usa7E3n5dhttcayTqFRm/van-gogh.jpeg"' \
  -i 'output_format="webp"' \
  -i 'output_quality=80' \
  -i 'negative_prompt=""' \
  -i 'number_of_images=1' \
  -i 'structure_depth_strength=1' \
  -i 'structure_denoising_strength=0.65'

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/fofr/style-transfer@sha256:f1023890703bc0a5a3a2c21b5e498833be5f6ef6e70e9daf6b9b3a4fd8309cf0
curl -s -X POST \
  -H "Content-Type: application/json" \
  -d $'{
    "input": {
      "model": "fast",
      "width": 1024,
      "height": 1024,
      "prompt": "An astronaut riding a unicorn",
      "style_image": "https://replicate.delivery/pbxt/KlTqluRakBzt7N5mm1WExEQCc4J3usa7E3n5dhttcayTqFRm/van-gogh.jpeg",
      "output_format": "webp",
      "output_quality": 80,
      "negative_prompt": "",
      "number_of_images": 1,
      "structure_depth_strength": 1,
      "structure_denoising_strength": 0.65
    }
  }' \
  http://localhost:5000/predictions

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

Output

{
  "completed_at": "2024-11-28T11:15:29.481256Z",
  "created_at": "2024-11-28T11:15:23.572000Z",
  "data_removed": false,
  "error": null,
  "id": "t71z6mcv6hrma0cke9xrtyzm1w",
  "input": {
    "model": "fast",
    "width": 1024,
    "height": 1024,
    "prompt": "An astronaut riding a unicorn",
    "style_image": "https://replicate.delivery/pbxt/KlTqluRakBzt7N5mm1WExEQCc4J3usa7E3n5dhttcayTqFRm/van-gogh.jpeg",
    "output_format": "webp",
    "output_quality": 80,
    "negative_prompt": "",
    "number_of_images": 1,
    "structure_depth_strength": 1,
    "structure_denoising_strength": 0.65
  },
  "logs": "Random seed set to: 1640868803\nChecking weights\nIncluding weights for IPAdapter preset: PLUS (high strength)\n✅ ip-adapter-plus_sdxl_vit-h.safetensors\n✅ CLIP-ViT-H-14-laion2B-s32B-b79K.safetensors\n✅ dreamshaperXL_lightningDPMSDE.safetensors\n====================================\nRunning workflow\ngot prompt\nExecuting node 2, title: Load Checkpoint, class type: CheckpointLoaderSimple\nmodel_type EPS\nUsing pytorch attention in VAE\nUsing pytorch attention in VAE\nclip missing: ['clip_l.logit_scale', 'clip_l.transformer.text_projection.weight']\nloaded straight to GPU\nRequested to load SDXL\nLoading 1 new model\nExecuting node 1, title: IPAdapter Unified Loader, class type: IPAdapterUnifiedLoader\n\u001b[33mINFO: Clip Vision model loaded from /src/ComfyUI/models/clip_vision/CLIP-ViT-H-14-laion2B-s32B-b79K.safetensors\u001b[0m\n\u001b[33mINFO: IPAdapter model loaded from /src/ComfyUI/models/ipadapter/ip-adapter-plus_sdxl_vit-h.safetensors\u001b[0m\nExecuting node 5, title: Load Image, class type: LoadImage\n\u001b[33mINFO: the IPAdapter reference image is not a square, CLIPImageProcessor will resize and crop it at the center. If the main focus of the picture is not in the middle the result might not be what you are expecting.\u001b[0m\nExecuting node 4, title: IPAdapter, class type: IPAdapter\nRequested to load CLIPVisionModelProjection\nLoading 1 new model\nExecuting node 6, title: CLIP Text Encode (Prompt), class type: CLIPTextEncode\nRequested to load SDXLClipModel\nLoading 1 new model\nExecuting node 7, title: CLIP Text Encode (Prompt), class type: CLIPTextEncode\nExecuting node 10, title: Empty Latent Image, class type: EmptyLatentImage\nExecuting node 3, title: KSampler, class type: KSampler\nRequested to load SDXL\nLoading 1 new model\n  0%|          | 0/4 [00:00<?, ?it/s]/root/.pyenv/versions/3.10.6/lib/python3.10/site-packages/torchsde/_brownian/brownian_interval.py:608: UserWarning: Should have tb<=t1 but got tb=14.614644050598145 and t1=14.614643.\nwarnings.warn(f\"Should have {tb_name}<=t1 but got {tb_name}={tb} and t1={self._end}.\")\n 25%|██▌       | 1/4 [00:00<00:01,  2.99it/s]\n 50%|█████     | 2/4 [00:00<00:00,  3.64it/s]\n 75%|███████▌  | 3/4 [00:00<00:00,  3.94it/s]\n100%|██████████| 4/4 [00:00<00:00,  5.05it/s]\n100%|██████████| 4/4 [00:00<00:00,  4.40it/s]\nRequested to load AutoencoderKL\nLoading 1 new model\nExecuting node 8, title: VAE Decode, class type: VAEDecode\nExecuting node 9, title: Save Image, class type: SaveImage\nPrompt executed in 5.15 seconds\noutputs:  {'9': {'images': [{'filename': 'ComfyUI_00001_.png', 'subfolder': '', 'type': 'output'}]}}\n====================================\nComfyUI_00001_.png",
  "metrics": {
    "predict_time": 5.8975421279999996,
    "total_time": 5.909256
  },
  "output": [
    "https://replicate.delivery/xezq/OgPfUK3cTWSYF6ZKr30ItkZClxuuf3INMygNv48O47wRLg1TA/ComfyUI_00001_.webp"
  ],
  "started_at": "2024-11-28T11:15:23.583714Z",
  "status": "succeeded",
  "urls": {
    "stream": "https://stream.replicate.com/v1/files/bcwr-ylw2cq4qmhgzi5wukon7sd7n2euktdfpoled6mp6tqvfknt5xnqq",
    "get": "https://api.replicate.com/v1/predictions/t71z6mcv6hrma0cke9xrtyzm1w",
    "cancel": "https://api.replicate.com/v1/predictions/t71z6mcv6hrma0cke9xrtyzm1w/cancel"
  },
  "version": "f1023890703bc0a5a3a2c21b5e498833be5f6ef6e70e9daf6b9b3a4fd8309cf0"
}

Generated in

5.9 seconds

Tweak it ShareReport View full prediction

Random seed set to: 1640868803
Checking weights
Including weights for IPAdapter preset: PLUS (high strength)
✅ ip-adapter-plus_sdxl_vit-h.safetensors
✅ CLIP-ViT-H-14-laion2B-s32B-b79K.safetensors
✅ dreamshaperXL_lightningDPMSDE.safetensors
====================================
Running workflow
got prompt
Executing node 2, title: Load Checkpoint, class type: CheckpointLoaderSimple
model_type EPS
Using pytorch attention in VAE
Using pytorch attention in VAE
clip missing: ['clip_l.logit_scale', 'clip_l.transformer.text_projection.weight']
loaded straight to GPU
Requested to load SDXL
Loading 1 new model
Executing node 1, title: IPAdapter Unified Loader, class type: IPAdapterUnifiedLoader
INFO: Clip Vision model loaded from /src/ComfyUI/models/clip_vision/CLIP-ViT-H-14-laion2B-s32B-b79K.safetensors
INFO: IPAdapter model loaded from /src/ComfyUI/models/ipadapter/ip-adapter-plus_sdxl_vit-h.safetensors
Executing node 5, title: Load Image, class type: LoadImage
INFO: the IPAdapter reference image is not a square, CLIPImageProcessor will resize and crop it at the center. If the main focus of the picture is not in the middle the result might not be what you are expecting.
Executing node 4, title: IPAdapter, class type: IPAdapter
Requested to load CLIPVisionModelProjection
Loading 1 new model
Executing node 6, title: CLIP Text Encode (Prompt), class type: CLIPTextEncode
Requested to load SDXLClipModel
Loading 1 new model
Executing node 7, title: CLIP Text Encode (Prompt), class type: CLIPTextEncode
Executing node 10, title: Empty Latent Image, class type: EmptyLatentImage
Executing node 3, title: KSampler, class type: KSampler
Requested to load SDXL
Loading 1 new model
  0%|          | 0/4 [00:00<?, ?it/s]/root/.pyenv/versions/3.10.6/lib/python3.10/site-packages/torchsde/_brownian/brownian_interval.py:608: UserWarning: Should have tb<=t1 but got tb=14.614644050598145 and t1=14.614643.
warnings.warn(f"Should have {tb_name}<=t1 but got {tb_name}={tb} and t1={self._end}.")
 25%|██▌       | 1/4 [00:00<00:01,  2.99it/s]
 50%|█████     | 2/4 [00:00<00:00,  3.64it/s]
 75%|███████▌  | 3/4 [00:00<00:00,  3.94it/s]
100%|██████████| 4/4 [00:00<00:00,  5.05it/s]
100%|██████████| 4/4 [00:00<00:00,  4.40it/s]
Requested to load AutoencoderKL
Loading 1 new model
Executing node 8, title: VAE Decode, class type: VAEDecode
Executing node 9, title: Save Image, class type: SaveImage
Prompt executed in 5.15 seconds
outputs:  {'9': {'images': [{'filename': 'ComfyUI_00001_.png', 'subfolder': '', 'type': 'output'}]}}
====================================
ComfyUI_00001_.png

Examples

View more examples

Run time and cost

This model costs approximately $0.0064 to run on Replicate, or 156 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 7 seconds.

Readme

Follow me on X: @fofrAI