lastmover/purr2 – Run with an API on Replicate

lastmover / purr2

Model trained to make PURR artwork

Warm

Public
129 runs
H100
GitHub

Run with an API

Playground API Examples README Versions

Input

prompt

*string

Shift + Return to add a new line

a digital illustration of PURR reading a book in a cozy librarya digital illustration of PURR reading a book in a cozy library

Prompt for generated image. If you include the `trigger_word` used in the training process you are more likely to activate the trained object, style, or concept in the resulting image.

image

file

Input image for image to image or inpainting mode. If provided, aspect_ratio, width, and height inputs are ignored.

mask

file

Image mask for image inpainting mode. If provided, aspect_ratio, width, and height inputs are ignored.

aspect_ratio

string

Aspect ratio for the generated image. If custom is selected, uses height and width below & will run in bf16 mode

Default: "1:1"

height

integer

(minimum: 256, maximum: 1440)

Height of generated image. Only works if `aspect_ratio` is set to custom. Will be rounded to nearest multiple of 16. Incompatible with fast generation

width

integer

(minimum: 256, maximum: 1440)

Width of generated image. Only works if `aspect_ratio` is set to custom. Will be rounded to nearest multiple of 16. Incompatible with fast generation

prompt_strength

number

(minimum: 0, maximum: 1)

Prompt strength when using img2img. 1.0 corresponds to full destruction of information in image

Default: 0.8

model

string

Which model to run inference with. The dev model performs best with around 28 inference steps but the schnell model only needs 4 steps.

Default: "dev"

num_outputs

integer

(minimum: 1, maximum: 4)

Number of outputs to generate

Default: 1

num_inference_steps

integer

(minimum: 1, maximum: 50)

Number of denoising steps. More steps can give more detailed images, but take longer.

Default: 28

guidance_scale

number

(minimum: 0, maximum: 10)

Guidance scale for the diffusion process. Lower values can give more realistic images. Good values to try are 2, 2.5, 3 and 3.5

Default: 3

seed

integer

Random seed. Set for reproducible generation

output_format

string

Format of the output images

Default: "webp"

output_quality

integer

(minimum: 0, maximum: 100)

Quality when saving the output images, from 0 to 100. 100 is best quality, 0 is lowest quality. Not relevant for .png outputs

Default: 80

disable_safety_checker

boolean

Disable safety checker for generated images.

Default: false

go_fast

boolean

Run faster predictions with model optimized for speed (currently fp8 quantized); disable to run in original bf16

Default: false

megapixels

string

Approximate number of megapixels for generated image

Default: "1"

lora_scale

number

(minimum: -1, maximum: 3)

Determines how strongly the main LoRA should be applied. Sane results between 0 and 1 for base inference. For go_fast we apply a 1.5x multiplier to this value; we've generally seen good performance when scaling the base value by that amount. You may still need to experiment to find the best value for your particular lora.

Default: 1

extra_lora

string

Shift + Return to add a new line

Load LoRA weights. Supports Replicate models in the format <owner>/<username> or <owner>/<username>/<version>, HuggingFace URLs in the format huggingface.co/<owner>/<model-name>, CivitAI URLs in the format civitai.com/models/<id>[/<model-name>], or arbitrary .safetensors URLs from the Internet. For example, 'fofr/flux-pixar-cars'

extra_lora_scale

number

(minimum: -1, maximum: 3)

Determines how strongly the extra LoRA should be applied. Sane results between 0 and 1 for base inference. For go_fast we apply a 1.5x multiplier to this value; we've generally seen good performance when scaling the base value by that amount. You may still need to experiment to find the best value for your particular lora.

Default: 1

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

npx create-replicate --model=lastmover/purr2

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 lastmover/purr2 using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "lastmover/purr2:19e5239358c8ecbc81acad82cd146ba16b5c6f4b52375cd170db578ece68b0f5",
  {
    input: {
      model: "dev",
      prompt: "a digital illustration of PURR reading a book in a cozy library",
      go_fast: false,
      lora_scale: 1,
      megapixels: "1",
      num_outputs: 4,
      aspect_ratio: "1:1",
      output_format: "webp",
      guidance_scale: 3,
      output_quality: 80,
      prompt_strength: 1,
      extra_lora_scale: 1,
      num_inference_steps: 28
    }
  }
);

// 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 lastmover/purr2 using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "lastmover/purr2:19e5239358c8ecbc81acad82cd146ba16b5c6f4b52375cd170db578ece68b0f5",
    input={
        "model": "dev",
        "prompt": "a digital illustration of PURR reading a book in a cozy library",
        "go_fast": False,
        "lora_scale": 1,
        "megapixels": "1",
        "num_outputs": 4,
        "aspect_ratio": "1:1",
        "output_format": "webp",
        "guidance_scale": 3,
        "output_quality": 80,
        "prompt_strength": 1,
        "extra_lora_scale": 1,
        "num_inference_steps": 28
    }
)
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 lastmover/purr2 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": "19e5239358c8ecbc81acad82cd146ba16b5c6f4b52375cd170db578ece68b0f5",
    "input": {
      "model": "dev",
      "prompt": "a digital illustration of PURR reading a book in a cozy library",
      "go_fast": false,
      "lora_scale": 1,
      "megapixels": "1",
      "num_outputs": 4,
      "aspect_ratio": "1:1",
      "output_format": "webp",
      "guidance_scale": 3,
      "output_quality": 80,
      "prompt_strength": 1,
      "extra_lora_scale": 1,
      "num_inference_steps": 28
    }
  }' \
  https://api.replicate.com/v1/predictions

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

Output

{
  "completed_at": "2025-01-28T01:47:35.589146Z",
  "created_at": "2025-01-28T01:47:07.694000Z",
  "data_removed": false,
  "error": null,
  "id": "kkhk9tt45srm80cmnabv871gec",
  "input": {
    "model": "dev",
    "prompt": "a digital illustration of PURR reading a book in a cozy library",
    "go_fast": false,
    "lora_scale": 1,
    "megapixels": "1",
    "num_outputs": 4,
    "aspect_ratio": "1:1",
    "output_format": "webp",
    "guidance_scale": 3,
    "output_quality": 80,
    "prompt_strength": 1,
    "extra_lora_scale": 1,
    "num_inference_steps": 28
  },
  "logs": "2025-01-28 01:47:08.531 | DEBUG    | fp8.lora_loading:apply_lora_to_model:574 - Extracting keys\n2025-01-28 01:47:08.532 | DEBUG    | fp8.lora_loading:apply_lora_to_model:581 - Keys extracted\nApplying LoRA:   0%|          | 0/304 [00:00<?, ?it/s]\nApplying LoRA:  89%|████████▉ | 270/304 [00:00<00:00, 2674.82it/s]\nApplying LoRA: 100%|██████████| 304/304 [00:00<00:00, 2606.04it/s]\n2025-01-28 01:47:08.649 | SUCCESS  | fp8.lora_loading:unload_loras:564 - LoRAs unloaded in 0.12s\nfree=28996837507072\nDownloading weights\n2025-01-28T01:47:08Z | INFO  | [ Initiating ] chunk_size=150M dest=/tmp/tmpkqjpahhz/weights url=https://replicate.delivery/xezq/nENtU0FT0XI8P5x2bRj3qRQQsMbM2kWgct4oGbvB49UjXXCF/trained_model.tar\n2025-01-28T01:47:11Z | INFO  | [ Complete ] dest=/tmp/tmpkqjpahhz/weights size=\"215 MB\" total_elapsed=2.556s url=https://replicate.delivery/xezq/nENtU0FT0XI8P5x2bRj3qRQQsMbM2kWgct4oGbvB49UjXXCF/trained_model.tar\nDownloaded weights in 2.58s\n2025-01-28 01:47:11.233 | INFO     | fp8.lora_loading:convert_lora_weights:498 - Loading LoRA weights for /src/weights-cache/7a19394bea9acc6e\n2025-01-28 01:47:11.314 | INFO     | fp8.lora_loading:convert_lora_weights:519 - LoRA weights loaded\n2025-01-28 01:47:11.315 | DEBUG    | fp8.lora_loading:apply_lora_to_model:574 - Extracting keys\n2025-01-28 01:47:11.315 | DEBUG    | fp8.lora_loading:apply_lora_to_model:581 - Keys extracted\nApplying LoRA:   0%|          | 0/304 [00:00<?, ?it/s]\nApplying LoRA:  92%|█████████▏| 281/304 [00:00<00:00, 2803.59it/s]\nApplying LoRA: 100%|██████████| 304/304 [00:00<00:00, 2752.52it/s]\n2025-01-28 01:47:11.426 | SUCCESS  | fp8.lora_loading:load_lora:539 - LoRA applied in 0.19s\nUsing seed: 3892\n0it [00:00, ?it/s]\n1it [00:00,  8.36it/s]\n2it [00:00,  5.86it/s]\n3it [00:00,  5.34it/s]\n4it [00:00,  5.12it/s]\n5it [00:00,  4.99it/s]\n6it [00:01,  4.91it/s]\n7it [00:01,  4.88it/s]\n8it [00:01,  4.85it/s]\n9it [00:01,  4.84it/s]\n10it [00:01,  4.82it/s]\n11it [00:02,  4.81it/s]\n12it [00:02,  4.81it/s]\n13it [00:02,  4.81it/s]\n14it [00:02,  4.82it/s]\n15it [00:03,  4.81it/s]\n16it [00:03,  4.81it/s]\n17it [00:03,  4.80it/s]\n18it [00:03,  4.80it/s]\n19it [00:03,  4.81it/s]\n20it [00:04,  4.81it/s]\n21it [00:04,  4.80it/s]\n22it [00:04,  4.79it/s]\n23it [00:04,  4.79it/s]\n24it [00:04,  4.79it/s]\n25it [00:05,  4.79it/s]\n26it [00:05,  4.80it/s]\n27it [00:05,  4.79it/s]\n28it [00:05,  4.80it/s]\n28it [00:05,  4.88it/s]\n0it [00:00, ?it/s]\n1it [00:00,  4.85it/s]\n2it [00:00,  4.82it/s]\n3it [00:00,  4.81it/s]\n4it [00:00,  4.80it/s]\n5it [00:01,  4.79it/s]\n6it [00:01,  4.79it/s]\n7it [00:01,  4.78it/s]\n8it [00:01,  4.78it/s]\n9it [00:01,  4.78it/s]\n10it [00:02,  4.78it/s]\n11it [00:02,  4.78it/s]\n12it [00:02,  4.77it/s]\n13it [00:02,  4.77it/s]\n14it [00:02,  4.76it/s]\n15it [00:03,  4.77it/s]\n16it [00:03,  4.77it/s]\n17it [00:03,  4.76it/s]\n18it [00:03,  4.76it/s]\n19it [00:03,  4.76it/s]\n20it [00:04,  4.77it/s]\n21it [00:04,  4.79it/s]\n22it [00:04,  4.78it/s]\n23it [00:04,  4.77it/s]\n24it [00:05,  4.78it/s]\n25it [00:05,  4.77it/s]\n26it [00:05,  4.77it/s]\n27it [00:05,  4.78it/s]\n28it [00:05,  4.79it/s]\n28it [00:05,  4.78it/s]\n0it [00:00, ?it/s]\n1it [00:00,  4.82it/s]\n2it [00:00,  4.79it/s]\n3it [00:00,  4.78it/s]\n4it [00:00,  4.78it/s]\n5it [00:01,  4.79it/s]\n6it [00:01,  4.79it/s]\n7it [00:01,  4.79it/s]\n8it [00:01,  4.79it/s]\n9it [00:01,  4.79it/s]\n10it [00:02,  4.79it/s]\n11it [00:02,  4.79it/s]\n12it [00:02,  4.79it/s]\n13it [00:02,  4.79it/s]\n14it [00:02,  4.80it/s]\n15it [00:03,  4.79it/s]\n16it [00:03,  4.79it/s]\n17it [00:03,  4.78it/s]\n18it [00:03,  4.78it/s]\n19it [00:03,  4.78it/s]\n20it [00:04,  4.78it/s]\n21it [00:04,  4.78it/s]\n22it [00:04,  4.79it/s]\n23it [00:04,  4.79it/s]\n24it [00:05,  4.79it/s]\n25it [00:05,  4.78it/s]\n26it [00:05,  4.78it/s]\n27it [00:05,  4.78it/s]\n28it [00:05,  4.77it/s]\n28it [00:05,  4.79it/s]\n0it [00:00, ?it/s]\n1it [00:00,  4.83it/s]\n2it [00:00,  4.81it/s]\n3it [00:00,  4.81it/s]\n4it [00:00,  4.81it/s]\n5it [00:01,  4.81it/s]\n6it [00:01,  4.79it/s]\n7it [00:01,  4.79it/s]\n8it [00:01,  4.78it/s]\n9it [00:01,  4.79it/s]\n10it [00:02,  4.79it/s]\n11it [00:02,  4.79it/s]\n12it [00:02,  4.79it/s]\n13it [00:02,  4.78it/s]\n14it [00:02,  4.77it/s]\n15it [00:03,  4.77it/s]\n16it [00:03,  4.78it/s]\n17it [00:03,  4.78it/s]\n18it [00:03,  4.78it/s]\n19it [00:03,  4.78it/s]\n20it [00:04,  4.78it/s]\n21it [00:04,  4.78it/s]\n22it [00:04,  4.78it/s]\n23it [00:04,  4.78it/s]\n24it [00:05,  4.78it/s]\n25it [00:05,  4.77it/s]\n26it [00:05,  4.78it/s]\n27it [00:05,  4.79it/s]\n28it [00:05,  4.79it/s]\n28it [00:05,  4.79it/s]\nTotal safe images: 4 out of 4",
  "metrics": {
    "predict_time": 27.056352393,
    "total_time": 27.895146
  },
  "output": [
    "https://replicate.delivery/xezq/L8idWx36Or7TCJJl4lpJyYR52DL6F5hansBLhrpkxe7bSvEKA/out-0.webp",
    "https://replicate.delivery/xezq/E8kbFzi5n4JADN5UhQAFyhGcAJ2M5jilXPgzJxMrkuyNpXCF/out-1.webp",
    "https://replicate.delivery/xezq/lCl11DVVs4JAH1KUOG1Rn0ppx7r7KEyhPldGn3Ftu3ebSvEKA/out-2.webp",
    "https://replicate.delivery/xezq/oHMI57Hmzw6hFJjAybf3sSRMtvLfwfCkeLTbkjvU8mIfm0LhC/out-3.webp"
  ],
  "started_at": "2025-01-28T01:47:08.532793Z",
  "status": "succeeded",
  "urls": {
    "stream": "https://stream.replicate.com/v1/files/bsvm-pe3nbrvn4voalo65t4dl3pgfd4a26une6mtj5c6kimw6wztf64ga",
    "get": "https://api.replicate.com/v1/predictions/kkhk9tt45srm80cmnabv871gec",
    "cancel": "https://api.replicate.com/v1/predictions/kkhk9tt45srm80cmnabv871gec/cancel"
  },
  "version": "19e5239358c8ecbc81acad82cd146ba16b5c6f4b52375cd170db578ece68b0f5"
}

Generated in

27.1 seconds

Tweak it ShareReport

2025-01-28 01:47:08.531 | DEBUG    | fp8.lora_loading:apply_lora_to_model:574 - Extracting keys
2025-01-28 01:47:08.532 | DEBUG    | fp8.lora_loading:apply_lora_to_model:581 - Keys extracted
Applying LoRA:   0%|          | 0/304 [00:00<?, ?it/s]
Applying LoRA:  89%|████████▉ | 270/304 [00:00<00:00, 2674.82it/s]
Applying LoRA: 100%|██████████| 304/304 [00:00<00:00, 2606.04it/s]
2025-01-28 01:47:08.649 | SUCCESS  | fp8.lora_loading:unload_loras:564 - LoRAs unloaded in 0.12s
free=28996837507072
Downloading weights
2025-01-28T01:47:08Z | INFO  | [ Initiating ] chunk_size=150M dest=/tmp/tmpkqjpahhz/weights url=https://replicate.delivery/xezq/nENtU0FT0XI8P5x2bRj3qRQQsMbM2kWgct4oGbvB49UjXXCF/trained_model.tar
2025-01-28T01:47:11Z | INFO  | [ Complete ] dest=/tmp/tmpkqjpahhz/weights size="215 MB" total_elapsed=2.556s url=https://replicate.delivery/xezq/nENtU0FT0XI8P5x2bRj3qRQQsMbM2kWgct4oGbvB49UjXXCF/trained_model.tar
Downloaded weights in 2.58s
2025-01-28 01:47:11.233 | INFO     | fp8.lora_loading:convert_lora_weights:498 - Loading LoRA weights for /src/weights-cache/7a19394bea9acc6e
2025-01-28 01:47:11.314 | INFO     | fp8.lora_loading:convert_lora_weights:519 - LoRA weights loaded
2025-01-28 01:47:11.315 | DEBUG    | fp8.lora_loading:apply_lora_to_model:574 - Extracting keys
2025-01-28 01:47:11.315 | DEBUG    | fp8.lora_loading:apply_lora_to_model:581 - Keys extracted
Applying LoRA:   0%|          | 0/304 [00:00<?, ?it/s]
Applying LoRA:  92%|█████████▏| 281/304 [00:00<00:00, 2803.59it/s]
Applying LoRA: 100%|██████████| 304/304 [00:00<00:00, 2752.52it/s]
2025-01-28 01:47:11.426 | SUCCESS  | fp8.lora_loading:load_lora:539 - LoRA applied in 0.19s
Using seed: 3892
0it [00:00, ?it/s]
1it [00:00,  8.36it/s]
2it [00:00,  5.86it/s]
3it [00:00,  5.34it/s]
4it [00:00,  5.12it/s]
5it [00:00,  4.99it/s]
6it [00:01,  4.91it/s]
7it [00:01,  4.88it/s]
8it [00:01,  4.85it/s]
9it [00:01,  4.84it/s]
10it [00:01,  4.82it/s]
11it [00:02,  4.81it/s]
12it [00:02,  4.81it/s]
13it [00:02,  4.81it/s]
14it [00:02,  4.82it/s]
15it [00:03,  4.81it/s]
16it [00:03,  4.81it/s]
17it [00:03,  4.80it/s]
18it [00:03,  4.80it/s]
19it [00:03,  4.81it/s]
20it [00:04,  4.81it/s]
21it [00:04,  4.80it/s]
22it [00:04,  4.79it/s]
23it [00:04,  4.79it/s]
24it [00:04,  4.79it/s]
25it [00:05,  4.79it/s]
26it [00:05,  4.80it/s]
27it [00:05,  4.79it/s]
28it [00:05,  4.80it/s]
28it [00:05,  4.88it/s]
0it [00:00, ?it/s]
1it [00:00,  4.85it/s]
2it [00:00,  4.82it/s]
3it [00:00,  4.81it/s]
4it [00:00,  4.80it/s]
5it [00:01,  4.79it/s]
6it [00:01,  4.79it/s]
7it [00:01,  4.78it/s]
8it [00:01,  4.78it/s]
9it [00:01,  4.78it/s]
10it [00:02,  4.78it/s]
11it [00:02,  4.78it/s]
12it [00:02,  4.77it/s]
13it [00:02,  4.77it/s]
14it [00:02,  4.76it/s]
15it [00:03,  4.77it/s]
16it [00:03,  4.77it/s]
17it [00:03,  4.76it/s]
18it [00:03,  4.76it/s]
19it [00:03,  4.76it/s]
20it [00:04,  4.77it/s]
21it [00:04,  4.79it/s]
22it [00:04,  4.78it/s]
23it [00:04,  4.77it/s]
24it [00:05,  4.78it/s]
25it [00:05,  4.77it/s]
26it [00:05,  4.77it/s]
27it [00:05,  4.78it/s]
28it [00:05,  4.79it/s]
28it [00:05,  4.78it/s]
0it [00:00, ?it/s]
1it [00:00,  4.82it/s]
2it [00:00,  4.79it/s]
3it [00:00,  4.78it/s]
4it [00:00,  4.78it/s]
5it [00:01,  4.79it/s]
6it [00:01,  4.79it/s]
7it [00:01,  4.79it/s]
8it [00:01,  4.79it/s]
9it [00:01,  4.79it/s]
10it [00:02,  4.79it/s]
11it [00:02,  4.79it/s]
12it [00:02,  4.79it/s]
13it [00:02,  4.79it/s]
14it [00:02,  4.80it/s]
15it [00:03,  4.79it/s]
16it [00:03,  4.79it/s]
17it [00:03,  4.78it/s]
18it [00:03,  4.78it/s]
19it [00:03,  4.78it/s]
20it [00:04,  4.78it/s]
21it [00:04,  4.78it/s]
22it [00:04,  4.79it/s]
23it [00:04,  4.79it/s]
24it [00:05,  4.79it/s]
25it [00:05,  4.78it/s]
26it [00:05,  4.78it/s]
27it [00:05,  4.78it/s]
28it [00:05,  4.77it/s]
28it [00:05,  4.79it/s]
0it [00:00, ?it/s]
1it [00:00,  4.83it/s]
2it [00:00,  4.81it/s]
3it [00:00,  4.81it/s]
4it [00:00,  4.81it/s]
5it [00:01,  4.81it/s]
6it [00:01,  4.79it/s]
7it [00:01,  4.79it/s]
8it [00:01,  4.78it/s]
9it [00:01,  4.79it/s]
10it [00:02,  4.79it/s]
11it [00:02,  4.79it/s]
12it [00:02,  4.79it/s]
13it [00:02,  4.78it/s]
14it [00:02,  4.77it/s]
15it [00:03,  4.77it/s]
16it [00:03,  4.78it/s]
17it [00:03,  4.78it/s]
18it [00:03,  4.78it/s]
19it [00:03,  4.78it/s]
20it [00:04,  4.78it/s]
21it [00:04,  4.78it/s]
22it [00:04,  4.78it/s]
23it [00:04,  4.78it/s]
24it [00:05,  4.78it/s]
25it [00:05,  4.77it/s]
26it [00:05,  4.78it/s]
27it [00:05,  4.79it/s]
28it [00:05,  4.79it/s]
28it [00:05,  4.79it/s]
Total safe images: 4 out of 4

Examples

View more examples

Run time and cost

This model costs approximately $0.040 to run on Replicate, or 25 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 H100 GPU hardware. Predictions typically complete within 27 seconds.

Readme

Use “PURR” as the trigger word in your prompt
For best results, structure prompts like:
“A digital illustration of PURR stargazing through a telescope”
“A digital illustration of PURR sipping coffee at a café”
The model works best with some experimentation - try varying your prompts to get optimal results