fofr/pulid-base – Run with an API on Replicate

fofr / pulid-base

Use a face to make images. Uses SDXL fine-tuned checkpoints.

Cold

Public
162.4K runs
L40S
GitHub
Paper

Run with an API

Playground API Examples README Versions

Input

face_image

*file

The face image to use for the generation

prompt

string

Shift + Return to add a new line

You might need to include a gender in the prompt to get the desired result

Default: "A photo of a person"

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

checkpoint_model

string

Model to use for the generation

Default: "general - dreamshaperXL_alpha2Xl10"

face_style

string

Style of the face

Default: "high-fidelity"

number_of_images

integer

(minimum: 1, maximum: 10)

Number of images to generate

Default: 1

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/pulid-base

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

const output = await replicate.run(
  "fofr/pulid-base:65ea75658bf120abbbdacab07e89e78a74a6a1b1f504349f4c4e3b01a655ee7a",
  {
    input: {
      width: 1024,
      height: 1024,
      prompt: "anime middle aged professional guy",
      face_image: "https://replicate.delivery/pbxt/KtD7T0zAQ4YU1vI4CptsLMSVzL239UWFHRcZ1DgWAhyVsfuh/guy.webp",
      face_style: "high-fidelity",
      output_format: "webp",
      output_quality: 80,
      negative_prompt: "boy",
      checkpoint_model: "animated - starlightXLAnimated_v3",
      number_of_images: 1
    }
  }
);
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/pulid-base using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "fofr/pulid-base:65ea75658bf120abbbdacab07e89e78a74a6a1b1f504349f4c4e3b01a655ee7a",
    input={
        "width": 1024,
        "height": 1024,
        "prompt": "anime middle aged professional guy",
        "face_image": "https://replicate.delivery/pbxt/KtD7T0zAQ4YU1vI4CptsLMSVzL239UWFHRcZ1DgWAhyVsfuh/guy.webp",
        "face_style": "high-fidelity",
        "output_format": "webp",
        "output_quality": 80,
        "negative_prompt": "boy",
        "checkpoint_model": "animated - starlightXLAnimated_v3",
        "number_of_images": 1
    }
)
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/pulid-base 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": "65ea75658bf120abbbdacab07e89e78a74a6a1b1f504349f4c4e3b01a655ee7a",
    "input": {
      "width": 1024,
      "height": 1024,
      "prompt": "anime middle aged professional guy",
      "face_image": "https://replicate.delivery/pbxt/KtD7T0zAQ4YU1vI4CptsLMSVzL239UWFHRcZ1DgWAhyVsfuh/guy.webp",
      "face_style": "high-fidelity",
      "output_format": "webp",
      "output_quality": 80,
      "negative_prompt": "boy",
      "checkpoint_model": "animated - starlightXLAnimated_v3",
      "number_of_images": 1
    }
  }' \
  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/pulid-base@sha256:65ea75658bf120abbbdacab07e89e78a74a6a1b1f504349f4c4e3b01a655ee7a \
  -i 'width=1024' \
  -i 'height=1024' \
  -i 'prompt="anime middle aged professional guy"' \
  -i 'face_image="https://replicate.delivery/pbxt/KtD7T0zAQ4YU1vI4CptsLMSVzL239UWFHRcZ1DgWAhyVsfuh/guy.webp"' \
  -i 'face_style="high-fidelity"' \
  -i 'output_format="webp"' \
  -i 'output_quality=80' \
  -i 'negative_prompt="boy"' \
  -i 'checkpoint_model="animated - starlightXLAnimated_v3"' \
  -i 'number_of_images=1'

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/pulid-base@sha256:65ea75658bf120abbbdacab07e89e78a74a6a1b1f504349f4c4e3b01a655ee7a
curl -s -X POST \
  -H "Content-Type: application/json" \
  -d $'{
    "input": {
      "width": 1024,
      "height": 1024,
      "prompt": "anime middle aged professional guy",
      "face_image": "https://replicate.delivery/pbxt/KtD7T0zAQ4YU1vI4CptsLMSVzL239UWFHRcZ1DgWAhyVsfuh/guy.webp",
      "face_style": "high-fidelity",
      "output_format": "webp",
      "output_quality": 80,
      "negative_prompt": "boy",
      "checkpoint_model": "animated - starlightXLAnimated_v3",
      "number_of_images": 1
    }
  }' \
  http://localhost:5000/predictions

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

Output

{
  "completed_at": "2024-05-09T15:16:07.004979Z",
  "created_at": "2024-05-09T15:15:59.174000Z",
  "data_removed": false,
  "error": null,
  "id": "3jagewqc8srgp0cfbqp9434csg",
  "input": {
    "width": 1024,
    "height": 1024,
    "prompt": "anime middle aged professional guy",
    "face_image": "https://replicate.delivery/pbxt/KtD7T0zAQ4YU1vI4CptsLMSVzL239UWFHRcZ1DgWAhyVsfuh/guy.webp",
    "face_style": "high-fidelity",
    "output_format": "webp",
    "output_quality": 80,
    "negative_prompt": "boy",
    "checkpoint_model": "animated - starlightXLAnimated_v3",
    "number_of_images": 1
  },
  "logs": "Random seed set to: 3813125192\nChecking inputs\n✅ /tmp/inputs/image.png\n====================================\nChecking weights\n✅ albedobaseXL_v21.safetensors\n✅ ip-adapter_pulid_sdxl_fp16.safetensors\n✅ pixlAnimeCartoonComic_v10.safetensors\n✅ ProteusV0.4-Lighting.safetensors\n✅ starlightXLAnimated_v3.safetensors\n✅ dreamshaperXL_alpha2Xl10.safetensors\n✅ models/antelopev2\n====================================\nRunning workflow\ngot prompt\nExecuting node 22, title: CLIP Text Encode (Prompt), class type: CLIPTextEncode\nExecuting node 3, title: KSampler, class type: KSampler\n  0%|          | 0/20 [00:00<?, ?it/s]\n  5%|▌         | 1/20 [00:00<00:04,  3.83it/s]\n 10%|█         | 2/20 [00:00<00:04,  3.77it/s]\n 15%|█▌        | 3/20 [00:00<00:04,  3.76it/s]\n 20%|██        | 4/20 [00:01<00:04,  3.77it/s]\n 25%|██▌       | 5/20 [00:01<00:03,  3.78it/s]\n 30%|███       | 6/20 [00:01<00:03,  3.77it/s]\n 35%|███▌      | 7/20 [00:01<00:03,  3.77it/s]\n 40%|████      | 8/20 [00:02<00:03,  3.77it/s]\n 45%|████▌     | 9/20 [00:02<00:02,  3.76it/s]\n 50%|█████     | 10/20 [00:02<00:02,  3.76it/s]\n 55%|█████▌    | 11/20 [00:02<00:02,  3.77it/s]\n 60%|██████    | 12/20 [00:03<00:02,  3.78it/s]\n 65%|██████▌   | 13/20 [00:03<00:01,  3.79it/s]\n 70%|███████   | 14/20 [00:03<00:01,  3.80it/s]\n 75%|███████▌  | 15/20 [00:03<00:01,  3.79it/s]\n 80%|████████  | 16/20 [00:04<00:01,  3.76it/s]\n 85%|████████▌ | 17/20 [00:04<00:00,  3.77it/s]\n 90%|█████████ | 18/20 [00:04<00:00,  3.79it/s]\n 95%|█████████▌| 19/20 [00:05<00:00,  3.87it/s]\n100%|██████████| 20/20 [00:05<00:00,  3.94it/s]\n100%|██████████| 20/20 [00:05<00:00,  3.81it/s]\nExecuting node 8, title: VAE Decode, class type: VAEDecode\nExecuting node 47, title: Save Image, class type: SaveImage\nPrompt executed in 5.66 seconds\noutputs:  {'47': {'images': [{'filename': 'ComfyUI_00001_.png', 'subfolder': '', 'type': 'output'}]}}\n====================================\nComfyUI_00001_.png",
  "metrics": {
    "predict_time": 7.792664,
    "total_time": 7.830979
  },
  "output": [
    "https://replicate.delivery/pbxt/jqivJUARfP3BV6XR9Qr8ktnXYSe5Ilagfstveh5bQSNYrmKLB/ComfyUI_00001_.webp"
  ],
  "started_at": "2024-05-09T15:15:59.212315Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/3jagewqc8srgp0cfbqp9434csg",
    "cancel": "https://api.replicate.com/v1/predictions/3jagewqc8srgp0cfbqp9434csg/cancel"
  },
  "version": "ff8800a68a85d1fed2bbba91c326ec82175b41a168e77c7c5b4653153ac9c3f1"
}

Generated in

7.8 seconds

Tweak it ShareReport

Random seed set to: 3813125192
Checking inputs
✅ /tmp/inputs/image.png
====================================
Checking weights
✅ albedobaseXL_v21.safetensors
✅ ip-adapter_pulid_sdxl_fp16.safetensors
✅ pixlAnimeCartoonComic_v10.safetensors
✅ ProteusV0.4-Lighting.safetensors
✅ starlightXLAnimated_v3.safetensors
✅ dreamshaperXL_alpha2Xl10.safetensors
✅ models/antelopev2
====================================
Running workflow
got prompt
Executing node 22, title: CLIP Text Encode (Prompt), class type: CLIPTextEncode
Executing node 3, title: KSampler, class type: KSampler
  0%|          | 0/20 [00:00<?, ?it/s]
  5%|▌         | 1/20 [00:00<00:04,  3.83it/s]
 10%|█         | 2/20 [00:00<00:04,  3.77it/s]
 15%|█▌        | 3/20 [00:00<00:04,  3.76it/s]
 20%|██        | 4/20 [00:01<00:04,  3.77it/s]
 25%|██▌       | 5/20 [00:01<00:03,  3.78it/s]
 30%|███       | 6/20 [00:01<00:03,  3.77it/s]
 35%|███▌      | 7/20 [00:01<00:03,  3.77it/s]
 40%|████      | 8/20 [00:02<00:03,  3.77it/s]
 45%|████▌     | 9/20 [00:02<00:02,  3.76it/s]
 50%|█████     | 10/20 [00:02<00:02,  3.76it/s]
 55%|█████▌    | 11/20 [00:02<00:02,  3.77it/s]
 60%|██████    | 12/20 [00:03<00:02,  3.78it/s]
 65%|██████▌   | 13/20 [00:03<00:01,  3.79it/s]
 70%|███████   | 14/20 [00:03<00:01,  3.80it/s]
 75%|███████▌  | 15/20 [00:03<00:01,  3.79it/s]
 80%|████████  | 16/20 [00:04<00:01,  3.76it/s]
 85%|████████▌ | 17/20 [00:04<00:00,  3.77it/s]
 90%|█████████ | 18/20 [00:04<00:00,  3.79it/s]
 95%|█████████▌| 19/20 [00:05<00:00,  3.87it/s]
100%|██████████| 20/20 [00:05<00:00,  3.94it/s]
100%|██████████| 20/20 [00:05<00:00,  3.81it/s]
Executing node 8, title: VAE Decode, class type: VAEDecode
Executing node 47, title: Save Image, class type: SaveImage
Prompt executed in 5.66 seconds
outputs:  {'47': {'images': [{'filename': 'ComfyUI_00001_.png', 'subfolder': '', 'type': 'output'}]}}
====================================
ComfyUI_00001_.png

This example was created by a different version, fofr/pulid-base:ff8800a6.

Examples

View more examples

Run time and cost

This model costs approximately $0.0067 to run on Replicate, or 149 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. The predict time for this model varies significantly based on the inputs.

Readme

Follow me on X @fofrAI

This model uses:

ComfyUI
PuLID_ComfyUI by @cubiq

It uses the following fine-tuned weights:

albedobaseXL_v21.safetensors
dreamshaperXL_alpha2Xl10.safetensors
starlightXLAnimated_v3.safetensors
pixlAnimeCartoonComic_v10.safetensors
rundiffusionXL_beta.safetensors
RealVisXL_V4.0.safetensors
sdxlUnstableDiffusers_nihilmania.safetensors
CinematicRedmond.safetensors

As part of its face detection is uses InsightFace’s antelopev2 model, which cannot be used commercially.