cjwbw/wavyfusion – Run with an API on Replicate

cjwbw / wavyfusion

dreambooth trained on a very diverse dataset ranging from photographs to paintings

Cold

Public
3.7K runs
T4

Run with an API

Playground API Examples README Versions

Input

prompt

string

Shift + Return to add a new line

Input prompt

Default: "blonde woman wa-vy style"

negative_prompt

string

Shift + Return to add a new line

Specify things to not see in the output

width

integer

Width of output image. Maximum size is 1024x768 or 768x1024 because of memory limits

Default: 512

height

integer

Height of output image. Maximum size is 1024x768 or 768x1024 because of memory limits

Default: 512

prompt_strength

number

Prompt strength when using init image. 1.0 corresponds to full destruction of information in init image

Default: 0.8

num_outputs

integer

(minimum: 1, maximum: 4)

Number of images to output.

Default: 1

num_inference_steps

integer

(minimum: 1, maximum: 500)

Number of denoising steps

Default: 50

guidance_scale

number

(minimum: 1, maximum: 20)

Scale for classifier-free guidance

Default: 7.5

scheduler

string

Choose a scheduler

Default: "K_EULER"

seed

integer

Random seed. Leave blank to randomize the seed

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

npx create-replicate --model=cjwbw/wavyfusion

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

const output = await replicate.run(
  "cjwbw/wavyfusion:3a38e1795ef77e3be3d6eb77fbafaeec79e67d13ee0025b36a93bb17e540efc9",
  {
    input: {
      width: 512,
      height: 512,
      prompt: "batman wa-vy style",
      scheduler: "DPMSolverMultistep",
      num_outputs: 1,
      guidance_scale: 7.5,
      prompt_strength: 0.8,
      num_inference_steps: 50
    }
  }
);

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

output = replicate.run(
    "cjwbw/wavyfusion:3a38e1795ef77e3be3d6eb77fbafaeec79e67d13ee0025b36a93bb17e540efc9",
    input={
        "width": 512,
        "height": 512,
        "prompt": "batman wa-vy style",
        "scheduler": "DPMSolverMultistep",
        "num_outputs": 1,
        "guidance_scale": 7.5,
        "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 cjwbw/wavyfusion 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": "cjwbw/wavyfusion:3a38e1795ef77e3be3d6eb77fbafaeec79e67d13ee0025b36a93bb17e540efc9",
    "input": {
      "width": 512,
      "height": 512,
      "prompt": "batman wa-vy style",
      "scheduler": "DPMSolverMultistep",
      "num_outputs": 1,
      "guidance_scale": 7.5,
      "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.

Output

{
  "completed_at": "2022-12-08T22:39:38.054324Z",
  "created_at": "2022-12-08T22:35:34.417392Z",
  "data_removed": false,
  "error": null,
  "id": "srdc4zbrofb4rnwhrxukyirza4",
  "input": {
    "width": 512,
    "height": 512,
    "prompt": "batman wa-vy style",
    "scheduler": "DPMSolverMultistep",
    "num_outputs": 1,
    "guidance_scale": 7.5,
    "prompt_strength": 0.8,
    "num_inference_steps": 50
  },
  "logs": "Using seed: 1205\n  0%|          | 0/50 [00:00<?, ?it/s]\n  2%|▏         | 1/50 [00:02<01:52,  2.30s/it]\n  4%|▍         | 2/50 [00:02<00:52,  1.09s/it]\n  6%|▌         | 3/50 [00:02<00:32,  1.43it/s]\n  8%|▊         | 4/50 [00:03<00:23,  1.94it/s]\n 10%|█         | 5/50 [00:03<00:18,  2.41it/s]\n 12%|█▏        | 6/50 [00:03<00:15,  2.82it/s]\n 14%|█▍        | 7/50 [00:03<00:13,  3.18it/s]\n 16%|█▌        | 8/50 [00:03<00:12,  3.46it/s]\n 18%|█▊        | 9/50 [00:04<00:11,  3.66it/s]\n 20%|██        | 10/50 [00:04<00:10,  3.81it/s]\n 22%|██▏       | 11/50 [00:04<00:09,  3.95it/s]\n 24%|██▍       | 12/50 [00:04<00:09,  4.04it/s]\n 26%|██▌       | 13/50 [00:05<00:09,  4.10it/s]\n 28%|██▊       | 14/50 [00:05<00:08,  4.14it/s]\n 30%|███       | 15/50 [00:05<00:08,  4.16it/s]\n 32%|███▏      | 16/50 [00:05<00:08,  4.19it/s]\n 34%|███▍      | 17/50 [00:06<00:07,  4.21it/s]\n 36%|███▌      | 18/50 [00:06<00:07,  4.21it/s]\n 38%|███▊      | 19/50 [00:06<00:07,  4.22it/s]\n 40%|████      | 20/50 [00:06<00:07,  4.23it/s]\n 42%|████▏     | 21/50 [00:07<00:06,  4.22it/s]\n 44%|████▍     | 22/50 [00:07<00:06,  4.22it/s]\n 46%|████▌     | 23/50 [00:07<00:06,  4.23it/s]\n 48%|████▊     | 24/50 [00:07<00:06,  4.22it/s]\n 50%|█████     | 25/50 [00:07<00:05,  4.23it/s]\n 52%|█████▏    | 26/50 [00:08<00:05,  4.22it/s]\n 54%|█████▍    | 27/50 [00:08<00:05,  4.23it/s]\n 56%|█████▌    | 28/50 [00:08<00:05,  4.23it/s]\n 58%|█████▊    | 29/50 [00:08<00:04,  4.24it/s]\n 60%|██████    | 30/50 [00:09<00:04,  4.23it/s]\n 62%|██████▏   | 31/50 [00:09<00:04,  4.22it/s]\n 64%|██████▍   | 32/50 [00:09<00:04,  4.24it/s]\n 66%|██████▌   | 33/50 [00:09<00:04,  4.22it/s]\n 68%|██████▊   | 34/50 [00:10<00:03,  4.23it/s]\n 70%|███████   | 35/50 [00:10<00:03,  4.24it/s]\n 72%|███████▏  | 36/50 [00:10<00:03,  4.23it/s]\n 74%|███████▍  | 37/50 [00:10<00:03,  4.23it/s]\n 76%|███████▌  | 38/50 [00:11<00:02,  4.23it/s]\n 78%|███████▊  | 39/50 [00:11<00:02,  4.22it/s]\n 80%|████████  | 40/50 [00:11<00:02,  4.22it/s]\n 82%|████████▏ | 41/50 [00:11<00:02,  4.22it/s]\n 84%|████████▍ | 42/50 [00:11<00:01,  4.22it/s]\n 86%|████████▌ | 43/50 [00:12<00:01,  4.21it/s]\n 88%|████████▊ | 44/50 [00:12<00:01,  4.21it/s]\n 90%|█████████ | 45/50 [00:12<00:01,  4.22it/s]\n 92%|█████████▏| 46/50 [00:12<00:00,  4.21it/s]\n 94%|█████████▍| 47/50 [00:13<00:00,  4.21it/s]\n 96%|█████████▌| 48/50 [00:13<00:00,  4.22it/s]\n 98%|█████████▊| 49/50 [00:13<00:00,  4.22it/s]\n100%|██████████| 50/50 [00:13<00:00,  4.22it/s]\n100%|██████████| 50/50 [00:13<00:00,  3.60it/s]",
  "metrics": {
    "predict_time": 16.287517,
    "total_time": 243.636932
  },
  "output": [
    "https://replicate.delivery/pbxt/X3QiWcxedlyrekButiNlJqVDdfRfcy9merjX6yfOSzAaqZACE/out-0.png"
  ],
  "started_at": "2022-12-08T22:39:21.766807Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/srdc4zbrofb4rnwhrxukyirza4",
    "cancel": "https://api.replicate.com/v1/predictions/srdc4zbrofb4rnwhrxukyirza4/cancel"
  },
  "version": "3a38e1795ef77e3be3d6eb77fbafaeec79e67d13ee0025b36a93bb17e540efc9"
}

Generated in

16.3 seconds

Tweak itReport

Using seed: 1205
  0%|          | 0/50 [00:00<?, ?it/s]
  2%|▏         | 1/50 [00:02<01:52,  2.30s/it]
  4%|▍         | 2/50 [00:02<00:52,  1.09s/it]
  6%|▌         | 3/50 [00:02<00:32,  1.43it/s]
  8%|▊         | 4/50 [00:03<00:23,  1.94it/s]
 10%|█         | 5/50 [00:03<00:18,  2.41it/s]
 12%|█▏        | 6/50 [00:03<00:15,  2.82it/s]
 14%|█▍        | 7/50 [00:03<00:13,  3.18it/s]
 16%|█▌        | 8/50 [00:03<00:12,  3.46it/s]
 18%|█▊        | 9/50 [00:04<00:11,  3.66it/s]
 20%|██        | 10/50 [00:04<00:10,  3.81it/s]
 22%|██▏       | 11/50 [00:04<00:09,  3.95it/s]
 24%|██▍       | 12/50 [00:04<00:09,  4.04it/s]
 26%|██▌       | 13/50 [00:05<00:09,  4.10it/s]
 28%|██▊       | 14/50 [00:05<00:08,  4.14it/s]
 30%|███       | 15/50 [00:05<00:08,  4.16it/s]
 32%|███▏      | 16/50 [00:05<00:08,  4.19it/s]
 34%|███▍      | 17/50 [00:06<00:07,  4.21it/s]
 36%|███▌      | 18/50 [00:06<00:07,  4.21it/s]
 38%|███▊      | 19/50 [00:06<00:07,  4.22it/s]
 40%|████      | 20/50 [00:06<00:07,  4.23it/s]
 42%|████▏     | 21/50 [00:07<00:06,  4.22it/s]
 44%|████▍     | 22/50 [00:07<00:06,  4.22it/s]
 46%|████▌     | 23/50 [00:07<00:06,  4.23it/s]
 48%|████▊     | 24/50 [00:07<00:06,  4.22it/s]
 50%|█████     | 25/50 [00:07<00:05,  4.23it/s]
 52%|█████▏    | 26/50 [00:08<00:05,  4.22it/s]
 54%|█████▍    | 27/50 [00:08<00:05,  4.23it/s]
 56%|█████▌    | 28/50 [00:08<00:05,  4.23it/s]
 58%|█████▊    | 29/50 [00:08<00:04,  4.24it/s]
 60%|██████    | 30/50 [00:09<00:04,  4.23it/s]
 62%|██████▏   | 31/50 [00:09<00:04,  4.22it/s]
 64%|██████▍   | 32/50 [00:09<00:04,  4.24it/s]
 66%|██████▌   | 33/50 [00:09<00:04,  4.22it/s]
 68%|██████▊   | 34/50 [00:10<00:03,  4.23it/s]
 70%|███████   | 35/50 [00:10<00:03,  4.24it/s]
 72%|███████▏  | 36/50 [00:10<00:03,  4.23it/s]
 74%|███████▍  | 37/50 [00:10<00:03,  4.23it/s]
 76%|███████▌  | 38/50 [00:11<00:02,  4.23it/s]
 78%|███████▊  | 39/50 [00:11<00:02,  4.22it/s]
 80%|████████  | 40/50 [00:11<00:02,  4.22it/s]
 82%|████████▏ | 41/50 [00:11<00:02,  4.22it/s]
 84%|████████▍ | 42/50 [00:11<00:01,  4.22it/s]
 86%|████████▌ | 43/50 [00:12<00:01,  4.21it/s]
 88%|████████▊ | 44/50 [00:12<00:01,  4.21it/s]
 90%|█████████ | 45/50 [00:12<00:01,  4.22it/s]
 92%|█████████▏| 46/50 [00:12<00:00,  4.21it/s]
 94%|█████████▍| 47/50 [00:13<00:00,  4.21it/s]
 96%|█████████▌| 48/50 [00:13<00:00,  4.22it/s]
 98%|█████████▊| 49/50 [00:13<00:00,  4.22it/s]
100%|██████████| 50/50 [00:13<00:00,  4.22it/s]
100%|██████████| 50/50 [00:13<00:00,  3.60it/s]

Examples

View more examples

Run time and cost

This model costs approximately $0.0057 to run on Replicate, or 175 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 T4 GPU hardware. Predictions typically complete within 26 seconds. The predict time for this model varies significantly based on the inputs.

Readme

weights from https://huggingface.co/wavymulder/wavyfusion

Wavyfusion Header CKPT DOWNLOAD LINK - This is a dreambooth trained on a very diverse dataset ranging from photographs to paintings. The goal was to make a varied, general purpose model for illustrated styles.

In your prompt, use the activation token: wa-vy style

We use wa-vy instead of wavy because ‘wavy style’ introduced unwanted oceans and wavy hair.

Trained from 1.5 with VAE.

There are a lot of cool styles you can achieve with this model. Please see this document where I share the parameters (prompt, sampler, seed, etc.) used for all example images.

Character Example Landscape Example

And here is an batch of 49 images (not cherrypicked) in both euler_a and DPM++ 2M Karras

Special thanks to Nitrosocke and Guizmus