lucataco/blueprint – Run with an API on Replicate

lucataco / blueprint

An SDXL fine-tune based on blueprints

Warm

Public
269 runs
L40S
SDXL fine-tune

Run with an API

Playground API Examples README Versions

Input

prompt

string

Shift + Return to add a new line

Input prompt

Default: "An astronaut riding a rainbow unicorn"

negative_prompt

string

Shift + Return to add a new line

Input 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: 1024

height

integer

Height of output image

Default: 1024

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: 50

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"

high_noise_frac

number

(minimum: 0, maximum: 1)

For expert_ensemble_refiner, the fraction of noise to use

Default: 0.8

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: true

lora_scale

number

(minimum: 0, maximum: 1)

LoRA additive scale. Only applicable on trained models.

Default: 0.6

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

Default: false

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

npx create-replicate --model=lucataco/blueprint

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

const output = await replicate.run(
  "lucataco/blueprint:3a5ba8bcee5ecd1d4676f0f099c319f916975db7b5d03cad51fe9ba6630de4e7",
  {
    input: {
      width: 1024,
      height: 1024,
      prompt: "A TOK blueprint of a museum",
      refine: "no_refiner",
      scheduler: "K_EULER",
      lora_scale: 0.6,
      num_outputs: 1,
      guidance_scale: 7.5,
      apply_watermark: true,
      high_noise_frac: 0.8,
      negative_prompt: "",
      prompt_strength: 0.8,
      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 lucataco/blueprint using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "lucataco/blueprint:3a5ba8bcee5ecd1d4676f0f099c319f916975db7b5d03cad51fe9ba6630de4e7",
    input={
        "width": 1024,
        "height": 1024,
        "prompt": "A TOK blueprint of a museum",
        "refine": "no_refiner",
        "scheduler": "K_EULER",
        "lora_scale": 0.6,
        "num_outputs": 1,
        "guidance_scale": 7.5,
        "apply_watermark": True,
        "high_noise_frac": 0.8,
        "negative_prompt": "",
        "prompt_strength": 0.8,
        "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 lucataco/blueprint 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": "3a5ba8bcee5ecd1d4676f0f099c319f916975db7b5d03cad51fe9ba6630de4e7",
    "input": {
      "width": 1024,
      "height": 1024,
      "prompt": "A TOK blueprint of a museum",
      "refine": "no_refiner",
      "scheduler": "K_EULER",
      "lora_scale": 0.6,
      "num_outputs": 1,
      "guidance_scale": 7.5,
      "apply_watermark": true,
      "high_noise_frac": 0.8,
      "negative_prompt": "",
      "prompt_strength": 0.8,
      "num_inference_steps": 50
    }
  }' \
  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/lucataco/blueprint@sha256:3a5ba8bcee5ecd1d4676f0f099c319f916975db7b5d03cad51fe9ba6630de4e7 \
  -i 'width=1024' \
  -i 'height=1024' \
  -i 'prompt="A TOK blueprint of a museum"' \
  -i 'refine="no_refiner"' \
  -i 'scheduler="K_EULER"' \
  -i 'lora_scale=0.6' \
  -i 'num_outputs=1' \
  -i 'guidance_scale=7.5' \
  -i 'apply_watermark=true' \
  -i 'high_noise_frac=0.8' \
  -i 'negative_prompt=""' \
  -i 'prompt_strength=0.8' \
  -i 'num_inference_steps=50'

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/lucataco/blueprint@sha256:3a5ba8bcee5ecd1d4676f0f099c319f916975db7b5d03cad51fe9ba6630de4e7
curl -s -X POST \
  -H "Content-Type: application/json" \
  -d $'{
    "input": {
      "width": 1024,
      "height": 1024,
      "prompt": "A TOK blueprint of a museum",
      "refine": "no_refiner",
      "scheduler": "K_EULER",
      "lora_scale": 0.6,
      "num_outputs": 1,
      "guidance_scale": 7.5,
      "apply_watermark": true,
      "high_noise_frac": 0.8,
      "negative_prompt": "",
      "prompt_strength": 0.8,
      "num_inference_steps": 50
    }
  }' \
  http://localhost:5000/predictions

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

Output

{
  "completed_at": "2023-09-04T19:19:41.527403Z",
  "created_at": "2023-09-04T19:19:25.792198Z",
  "data_removed": false,
  "error": null,
  "id": "7tuxedlbgpsallbbybv5gwv2bi",
  "input": {
    "width": 1024,
    "height": 1024,
    "prompt": "A TOK blueprint of a museum",
    "refine": "no_refiner",
    "scheduler": "K_EULER",
    "lora_scale": 0.6,
    "num_outputs": 1,
    "guidance_scale": 7.5,
    "apply_watermark": true,
    "high_noise_frac": 0.8,
    "prompt_strength": 0.8,
    "num_inference_steps": 50
  },
  "logs": "Using seed: 1992\nPrompt: A <s0><s1> blueprint of a museum\ntxt2img mode\n  0%|          | 0/50 [00:00<?, ?it/s]\n  2%|▏         | 1/50 [00:00<00:13,  3.72it/s]\n  4%|▍         | 2/50 [00:00<00:12,  3.70it/s]\n  6%|▌         | 3/50 [00:00<00:12,  3.69it/s]\n  8%|▊         | 4/50 [00:01<00:12,  3.68it/s]\n 10%|█         | 5/50 [00:01<00:12,  3.68it/s]\n 12%|█▏        | 6/50 [00:01<00:11,  3.68it/s]\n 14%|█▍        | 7/50 [00:01<00:11,  3.68it/s]\n 16%|█▌        | 8/50 [00:02<00:11,  3.67it/s]\n 18%|█▊        | 9/50 [00:02<00:11,  3.67it/s]\n 20%|██        | 10/50 [00:02<00:10,  3.67it/s]\n 22%|██▏       | 11/50 [00:02<00:10,  3.67it/s]\n 24%|██▍       | 12/50 [00:03<00:10,  3.67it/s]\n 26%|██▌       | 13/50 [00:03<00:10,  3.67it/s]\n 28%|██▊       | 14/50 [00:03<00:09,  3.67it/s]\n 30%|███       | 15/50 [00:04<00:09,  3.67it/s]\n 32%|███▏      | 16/50 [00:04<00:09,  3.67it/s]\n 34%|███▍      | 17/50 [00:04<00:08,  3.67it/s]\n 36%|███▌      | 18/50 [00:04<00:08,  3.67it/s]\n 38%|███▊      | 19/50 [00:05<00:08,  3.67it/s]\n 40%|████      | 20/50 [00:05<00:08,  3.67it/s]\n 42%|████▏     | 21/50 [00:05<00:07,  3.67it/s]\n 44%|████▍     | 22/50 [00:05<00:07,  3.66it/s]\n 46%|████▌     | 23/50 [00:06<00:07,  3.66it/s]\n 48%|████▊     | 24/50 [00:06<00:07,  3.66it/s]\n 50%|█████     | 25/50 [00:06<00:06,  3.66it/s]\n 52%|█████▏    | 26/50 [00:07<00:06,  3.66it/s]\n 54%|█████▍    | 27/50 [00:07<00:06,  3.66it/s]\n 56%|█████▌    | 28/50 [00:07<00:06,  3.66it/s]\n 58%|█████▊    | 29/50 [00:07<00:05,  3.66it/s]\n 60%|██████    | 30/50 [00:08<00:05,  3.66it/s]\n 62%|██████▏   | 31/50 [00:08<00:05,  3.66it/s]\n 64%|██████▍   | 32/50 [00:08<00:04,  3.67it/s]\n 66%|██████▌   | 33/50 [00:08<00:04,  3.67it/s]\n 68%|██████▊   | 34/50 [00:09<00:04,  3.67it/s]\n 70%|███████   | 35/50 [00:09<00:04,  3.68it/s]\n 72%|███████▏  | 36/50 [00:09<00:03,  3.68it/s]\n 74%|███████▍  | 37/50 [00:10<00:03,  3.68it/s]\n 76%|███████▌  | 38/50 [00:10<00:03,  3.64it/s]\n 78%|███████▊  | 39/50 [00:10<00:03,  3.65it/s]\n 80%|████████  | 40/50 [00:10<00:02,  3.66it/s]\n 82%|████████▏ | 41/50 [00:11<00:02,  3.66it/s]\n 84%|████████▍ | 42/50 [00:11<00:02,  3.67it/s]\n 86%|████████▌ | 43/50 [00:11<00:01,  3.68it/s]\n 88%|████████▊ | 44/50 [00:11<00:01,  3.68it/s]\n 90%|█████████ | 45/50 [00:12<00:01,  3.68it/s]\n 92%|█████████▏| 46/50 [00:12<00:01,  3.68it/s]\n 94%|█████████▍| 47/50 [00:12<00:00,  3.68it/s]\n 96%|█████████▌| 48/50 [00:13<00:00,  3.68it/s]\n 98%|█████████▊| 49/50 [00:13<00:00,  3.68it/s]\n100%|██████████| 50/50 [00:13<00:00,  3.68it/s]\n100%|██████████| 50/50 [00:13<00:00,  3.67it/s]",
  "metrics": {
    "predict_time": 15.725051,
    "total_time": 15.735205
  },
  "output": [
    "https://pbxt.replicate.delivery/LDKCSsbuPx50AtWx43O8YcQRQIKWrrQGzLrq97LyGyFzfegRA/out-0.png"
  ],
  "started_at": "2023-09-04T19:19:25.802352Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/7tuxedlbgpsallbbybv5gwv2bi",
    "cancel": "https://api.replicate.com/v1/predictions/7tuxedlbgpsallbbybv5gwv2bi/cancel"
  },
  "version": "3a5ba8bcee5ecd1d4676f0f099c319f916975db7b5d03cad51fe9ba6630de4e7"
}

Generated in

15.7 seconds

Tweak itReport

Using seed: 1992
Prompt: A <s0><s1> blueprint of a museum
txt2img mode
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:00<00:13,  3.72it/s]
  4%|▍         | 2/50 [00:00<00:12,  3.70it/s]
  6%|▌         | 3/50 [00:00<00:12,  3.69it/s]
  8%|▊         | 4/50 [00:01<00:12,  3.68it/s]
 10%|█         | 5/50 [00:01<00:12,  3.68it/s]
 12%|█▏        | 6/50 [00:01<00:11,  3.68it/s]
 14%|█▍        | 7/50 [00:01<00:11,  3.68it/s]
 16%|█▌        | 8/50 [00:02<00:11,  3.67it/s]
 18%|█▊        | 9/50 [00:02<00:11,  3.67it/s]
 20%|██        | 10/50 [00:02<00:10,  3.67it/s]
 22%|██▏       | 11/50 [00:02<00:10,  3.67it/s]
 24%|██▍       | 12/50 [00:03<00:10,  3.67it/s]
 26%|██▌       | 13/50 [00:03<00:10,  3.67it/s]
 28%|██▊       | 14/50 [00:03<00:09,  3.67it/s]
 30%|███       | 15/50 [00:04<00:09,  3.67it/s]
 32%|███▏      | 16/50 [00:04<00:09,  3.67it/s]
 34%|███▍      | 17/50 [00:04<00:08,  3.67it/s]
 36%|███▌      | 18/50 [00:04<00:08,  3.67it/s]
 38%|███▊      | 19/50 [00:05<00:08,  3.67it/s]
 40%|████      | 20/50 [00:05<00:08,  3.67it/s]
 42%|████▏     | 21/50 [00:05<00:07,  3.67it/s]
 44%|████▍     | 22/50 [00:05<00:07,  3.66it/s]
 46%|████▌     | 23/50 [00:06<00:07,  3.66it/s]
 48%|████▊     | 24/50 [00:06<00:07,  3.66it/s]
 50%|█████     | 25/50 [00:06<00:06,  3.66it/s]
 52%|█████▏    | 26/50 [00:07<00:06,  3.66it/s]
 54%|█████▍    | 27/50 [00:07<00:06,  3.66it/s]
 56%|█████▌    | 28/50 [00:07<00:06,  3.66it/s]
 58%|█████▊    | 29/50 [00:07<00:05,  3.66it/s]
 60%|██████    | 30/50 [00:08<00:05,  3.66it/s]
 62%|██████▏   | 31/50 [00:08<00:05,  3.66it/s]
 64%|██████▍   | 32/50 [00:08<00:04,  3.67it/s]
 66%|██████▌   | 33/50 [00:08<00:04,  3.67it/s]
 68%|██████▊   | 34/50 [00:09<00:04,  3.67it/s]
 70%|███████   | 35/50 [00:09<00:04,  3.68it/s]
 72%|███████▏  | 36/50 [00:09<00:03,  3.68it/s]
 74%|███████▍  | 37/50 [00:10<00:03,  3.68it/s]
 76%|███████▌  | 38/50 [00:10<00:03,  3.64it/s]
 78%|███████▊  | 39/50 [00:10<00:03,  3.65it/s]
 80%|████████  | 40/50 [00:10<00:02,  3.66it/s]
 82%|████████▏ | 41/50 [00:11<00:02,  3.66it/s]
 84%|████████▍ | 42/50 [00:11<00:02,  3.67it/s]
 86%|████████▌ | 43/50 [00:11<00:01,  3.68it/s]
 88%|████████▊ | 44/50 [00:11<00:01,  3.68it/s]
 90%|█████████ | 45/50 [00:12<00:01,  3.68it/s]
 92%|█████████▏| 46/50 [00:12<00:01,  3.68it/s]
 94%|█████████▍| 47/50 [00:12<00:00,  3.68it/s]
 96%|█████████▌| 48/50 [00:13<00:00,  3.68it/s]
 98%|█████████▊| 49/50 [00:13<00:00,  3.68it/s]
100%|██████████| 50/50 [00:13<00:00,  3.68it/s]
100%|██████████| 50/50 [00:13<00:00,  3.67it/s]

Examples

View more examples

Run time and cost

This model costs approximately $0.022 to run on Replicate, or 45 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 23 seconds. The predict time for this model varies significantly based on the inputs.

Readme

Attempt at creating an SDXL finetune on blueprints

Let me know if you like my work! @lucataco93