fofr/sdxl-reference – Run with an API on Replicate

fofr / sdxl-reference

Make variations of an image with SDXL

Cold

Public
357 runs
L40S
GitHub

Run with an API

Playground API Examples Train README Versions

Input

prompt

string

Shift + Return to add a new line

Input prompt

Default: ""

negative_prompt

string

Shift + Return to add a new line

Negative prompt

Default: ""

image

file

Reference image to make variations of

width

integer

Width of output image

Default: 768

height

integer

Height of output image

Default: 768

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

seed

integer

Random seed. Leave blank to randomize the seed

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

disable_safety_checker

boolean

Disable safety checker for generated images. This feature is only available through the API. See [https://replicate.com/docs/how-does-replicate-work#safety](https://replicate.com/docs/how-does-replicate-work#safety)

Default: false

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

npx create-replicate --model=fofr/sdxl-reference

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

const output = await replicate.run(
  "fofr/sdxl-reference:255c5260c64a99de3c76bb349d517e75ef59280d97cae56b1b4118010075d98b",
  {
    input: {
      image: "https://replicate.delivery/pbxt/Jry0X2SDwNDCWxVpn9ndrlj4RciPlOntsGui7PY3JZtu5fmM/output-0-1.png",
      width: 768,
      height: 768,
      prompt: "a cat yawning",
      scheduler: "K_EULER",
      num_outputs: 4,
      guidance_scale: 7.5,
      apply_watermark: false,
      negative_prompt: "",
      num_inference_steps: 50
    }
  }
);
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/sdxl-reference using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "fofr/sdxl-reference:255c5260c64a99de3c76bb349d517e75ef59280d97cae56b1b4118010075d98b",
    input={
        "image": "https://replicate.delivery/pbxt/Jry0X2SDwNDCWxVpn9ndrlj4RciPlOntsGui7PY3JZtu5fmM/output-0-1.png",
        "width": 768,
        "height": 768,
        "prompt": "a cat yawning",
        "scheduler": "K_EULER",
        "num_outputs": 4,
        "guidance_scale": 7.5,
        "apply_watermark": False,
        "negative_prompt": "",
        "num_inference_steps": 50
    }
)
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/sdxl-reference 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": "255c5260c64a99de3c76bb349d517e75ef59280d97cae56b1b4118010075d98b",
    "input": {
      "image": "https://replicate.delivery/pbxt/Jry0X2SDwNDCWxVpn9ndrlj4RciPlOntsGui7PY3JZtu5fmM/output-0-1.png",
      "width": 768,
      "height": 768,
      "prompt": "a cat yawning",
      "scheduler": "K_EULER",
      "num_outputs": 4,
      "guidance_scale": 7.5,
      "apply_watermark": false,
      "negative_prompt": "",
      "num_inference_steps": 50
    }
  }' \
  https://api.replicate.com/v1/predictions

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

Output

{
  "completed_at": "2023-11-13T11:07:26.073442Z",
  "created_at": "2023-11-13T11:06:26.654904Z",
  "data_removed": false,
  "error": null,
  "id": "jsiqxplbbhmdri7idhg2p2cgwu",
  "input": {
    "image": "https://replicate.delivery/pbxt/Jry0X2SDwNDCWxVpn9ndrlj4RciPlOntsGui7PY3JZtu5fmM/output-0-1.png",
    "width": 768,
    "height": 768,
    "prompt": "a cat yawning",
    "scheduler": "K_EULER",
    "num_outputs": 4,
    "guidance_scale": 7.5,
    "apply_watermark": false,
    "negative_prompt": "",
    "num_inference_steps": 50
  },
  "logs": "Using seed: 49968\nPrompt: a cat yawning\n  0%|          | 0/50 [00:00<?, ?it/s]\n  2%|▏         | 1/50 [00:01<00:52,  1.06s/it]\n  4%|▍         | 2/50 [00:02<00:51,  1.07s/it]\n  6%|▌         | 3/50 [00:03<00:50,  1.07s/it]\n  8%|▊         | 4/50 [00:04<00:49,  1.07s/it]\n 10%|█         | 5/50 [00:05<00:47,  1.07s/it]\n 12%|█▏        | 6/50 [00:06<00:46,  1.07s/it]\n 14%|█▍        | 7/50 [00:07<00:45,  1.07s/it]\n 16%|█▌        | 8/50 [00:08<00:44,  1.07s/it]\n 18%|█▊        | 9/50 [00:09<00:43,  1.07s/it]\n 20%|██        | 10/50 [00:10<00:42,  1.07s/it]\n 22%|██▏       | 11/50 [00:11<00:41,  1.07s/it]\n 24%|██▍       | 12/50 [00:12<00:40,  1.07s/it]\n 26%|██▌       | 13/50 [00:13<00:39,  1.07s/it]\n 28%|██▊       | 14/50 [00:14<00:38,  1.07s/it]\n 30%|███       | 15/50 [00:16<00:37,  1.07s/it]\n 32%|███▏      | 16/50 [00:17<00:36,  1.07s/it]\n 34%|███▍      | 17/50 [00:18<00:35,  1.07s/it]\n 36%|███▌      | 18/50 [00:19<00:34,  1.07s/it]\n 38%|███▊      | 19/50 [00:20<00:33,  1.07s/it]\n 40%|████      | 20/50 [00:21<00:32,  1.08s/it]\n 42%|████▏     | 21/50 [00:22<00:31,  1.08s/it]\n 44%|████▍     | 22/50 [00:23<00:30,  1.08s/it]\n 46%|████▌     | 23/50 [00:24<00:29,  1.08s/it]\n 48%|████▊     | 24/50 [00:25<00:28,  1.08s/it]\n 50%|█████     | 25/50 [00:26<00:26,  1.08s/it]\n 52%|█████▏    | 26/50 [00:27<00:25,  1.08s/it]\n 54%|█████▍    | 27/50 [00:28<00:24,  1.08s/it]\n 56%|█████▌    | 28/50 [00:30<00:23,  1.08s/it]\n 58%|█████▊    | 29/50 [00:31<00:22,  1.08s/it]\n 60%|██████    | 30/50 [00:32<00:21,  1.08s/it]\n 62%|██████▏   | 31/50 [00:33<00:20,  1.08s/it]\n 64%|██████▍   | 32/50 [00:34<00:19,  1.08s/it]\n 66%|██████▌   | 33/50 [00:35<00:18,  1.08s/it]\n 68%|██████▊   | 34/50 [00:36<00:17,  1.08s/it]\n 70%|███████   | 35/50 [00:37<00:16,  1.08s/it]\n 72%|███████▏  | 36/50 [00:38<00:15,  1.08s/it]\n 74%|███████▍  | 37/50 [00:39<00:14,  1.08s/it]\n 76%|███████▌  | 38/50 [00:40<00:13,  1.08s/it]\n 78%|███████▊  | 39/50 [00:41<00:11,  1.08s/it]\n 80%|████████  | 40/50 [00:43<00:10,  1.08s/it]\n 82%|████████▏ | 41/50 [00:44<00:09,  1.08s/it]\n 84%|████████▍ | 42/50 [00:45<00:08,  1.08s/it]\n 86%|████████▌ | 43/50 [00:46<00:07,  1.09s/it]\n 88%|████████▊ | 44/50 [00:47<00:06,  1.09s/it]\n 90%|█████████ | 45/50 [00:48<00:05,  1.09s/it]\n 92%|█████████▏| 46/50 [00:49<00:04,  1.09s/it]\n 94%|█████████▍| 47/50 [00:50<00:03,  1.09s/it]\n 96%|█████████▌| 48/50 [00:51<00:02,  1.09s/it]\n 98%|█████████▊| 49/50 [00:52<00:01,  1.09s/it]\n100%|██████████| 50/50 [00:53<00:00,  1.09s/it]\n100%|██████████| 50/50 [00:53<00:00,  1.08s/it]",
  "metrics": {
    "predict_time": 59.474061,
    "total_time": 59.418538
  },
  "output": [
    "https://replicate.delivery/pbxt/63GQh2TMiioaIxw3AtHLth11MIYKoLZtMLNgvwmHNw4a1e7IA/out-0.png",
    "https://replicate.delivery/pbxt/fzofLKIJ19qib0i8BJGMIFrM0Ycu0V4T21OIob7YxYfYr2vjA/out-1.png",
    "https://replicate.delivery/pbxt/ETibthKf60yfSkc6FeQSgBFdWtSdJBzdpctcfDtrAj63WtfOC/out-2.png",
    "https://replicate.delivery/pbxt/DEfR4Nu6AXS2GCBIbKJ5pvuK174abvTg9S9WkIBpkeQtV73RA/out-3.png"
  ],
  "started_at": "2023-11-13T11:06:26.599381Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/jsiqxplbbhmdri7idhg2p2cgwu",
    "cancel": "https://api.replicate.com/v1/predictions/jsiqxplbbhmdri7idhg2p2cgwu/cancel"
  },
  "version": "255c5260c64a99de3c76bb349d517e75ef59280d97cae56b1b4118010075d98b"
}

Generated in

59.5 seconds

Tweak itReport

Using seed: 49968
Prompt: a cat yawning
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:01<00:52,  1.06s/it]
  4%|▍         | 2/50 [00:02<00:51,  1.07s/it]
  6%|▌         | 3/50 [00:03<00:50,  1.07s/it]
  8%|▊         | 4/50 [00:04<00:49,  1.07s/it]
 10%|█         | 5/50 [00:05<00:47,  1.07s/it]
 12%|█▏        | 6/50 [00:06<00:46,  1.07s/it]
 14%|█▍        | 7/50 [00:07<00:45,  1.07s/it]
 16%|█▌        | 8/50 [00:08<00:44,  1.07s/it]
 18%|█▊        | 9/50 [00:09<00:43,  1.07s/it]
 20%|██        | 10/50 [00:10<00:42,  1.07s/it]
 22%|██▏       | 11/50 [00:11<00:41,  1.07s/it]
 24%|██▍       | 12/50 [00:12<00:40,  1.07s/it]
 26%|██▌       | 13/50 [00:13<00:39,  1.07s/it]
 28%|██▊       | 14/50 [00:14<00:38,  1.07s/it]
 30%|███       | 15/50 [00:16<00:37,  1.07s/it]
 32%|███▏      | 16/50 [00:17<00:36,  1.07s/it]
 34%|███▍      | 17/50 [00:18<00:35,  1.07s/it]
 36%|███▌      | 18/50 [00:19<00:34,  1.07s/it]
 38%|███▊      | 19/50 [00:20<00:33,  1.07s/it]
 40%|████      | 20/50 [00:21<00:32,  1.08s/it]
 42%|████▏     | 21/50 [00:22<00:31,  1.08s/it]
 44%|████▍     | 22/50 [00:23<00:30,  1.08s/it]
 46%|████▌     | 23/50 [00:24<00:29,  1.08s/it]
 48%|████▊     | 24/50 [00:25<00:28,  1.08s/it]
 50%|█████     | 25/50 [00:26<00:26,  1.08s/it]
 52%|█████▏    | 26/50 [00:27<00:25,  1.08s/it]
 54%|█████▍    | 27/50 [00:28<00:24,  1.08s/it]
 56%|█████▌    | 28/50 [00:30<00:23,  1.08s/it]
 58%|█████▊    | 29/50 [00:31<00:22,  1.08s/it]
 60%|██████    | 30/50 [00:32<00:21,  1.08s/it]
 62%|██████▏   | 31/50 [00:33<00:20,  1.08s/it]
 64%|██████▍   | 32/50 [00:34<00:19,  1.08s/it]
 66%|██████▌   | 33/50 [00:35<00:18,  1.08s/it]
 68%|██████▊   | 34/50 [00:36<00:17,  1.08s/it]
 70%|███████   | 35/50 [00:37<00:16,  1.08s/it]
 72%|███████▏  | 36/50 [00:38<00:15,  1.08s/it]
 74%|███████▍  | 37/50 [00:39<00:14,  1.08s/it]
 76%|███████▌  | 38/50 [00:40<00:13,  1.08s/it]
 78%|███████▊  | 39/50 [00:41<00:11,  1.08s/it]
 80%|████████  | 40/50 [00:43<00:10,  1.08s/it]
 82%|████████▏ | 41/50 [00:44<00:09,  1.08s/it]
 84%|████████▍ | 42/50 [00:45<00:08,  1.08s/it]
 86%|████████▌ | 43/50 [00:46<00:07,  1.09s/it]
 88%|████████▊ | 44/50 [00:47<00:06,  1.09s/it]
 90%|█████████ | 45/50 [00:48<00:05,  1.09s/it]
 92%|█████████▏| 46/50 [00:49<00:04,  1.09s/it]
 94%|█████████▍| 47/50 [00:50<00:03,  1.09s/it]
 96%|█████████▌| 48/50 [00:51<00:02,  1.09s/it]
 98%|█████████▊| 49/50 [00:52<00:01,  1.09s/it]
100%|██████████| 50/50 [00:53<00:00,  1.09s/it]
100%|██████████| 50/50 [00:53<00:00,  1.08s/it]

Examples

View more examples

Run time and cost

This model costs approximately $0.058 to run on Replicate, or 17 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 60 seconds. The predict time for this model varies significantly based on the inputs.

Readme

Based on the diffusers community pipeline: https://github.com/huggingface/diffusers/blob/main/examples/community/stable_diffusion_xl_reference.py