lambdal/text-to-pokemon – Run with an API on Replicate

lambdal / text-to-pokemon

Generate Pokémon from a text description

Cold

Public
7.9M runs
T4
GitHub
License

Run with an API

Playground API Examples README Versions

Input

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

npx create-replicate --model=lambdal/text-to-pokemon

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 lambdal/text-to-pokemon using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "lambdal/text-to-pokemon:ff6cc781634191dd3c49097a615d2fc01b0a8aae31c448e55039a04dcbf36bba",
  {
    input: {
      prompt: "Yoda",
      num_outputs: 1,
      guidance_scale: 7.5,
      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 lambdal/text-to-pokemon using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "lambdal/text-to-pokemon:ff6cc781634191dd3c49097a615d2fc01b0a8aae31c448e55039a04dcbf36bba",
    input={
        "prompt": "Yoda",
        "num_outputs": 1,
        "guidance_scale": 7.5,
        "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 lambdal/text-to-pokemon 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": "ff6cc781634191dd3c49097a615d2fc01b0a8aae31c448e55039a04dcbf36bba",
    "input": {
      "prompt": "Yoda",
      "num_outputs": 1,
      "guidance_scale": 7.5,
      "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/lambdal/text-to-pokemon@sha256:ff6cc781634191dd3c49097a615d2fc01b0a8aae31c448e55039a04dcbf36bba \
  -i 'prompt="Yoda"' \
  -i 'num_outputs=1' \
  -i 'guidance_scale=7.5' \
  -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/lambdal/text-to-pokemon@sha256:ff6cc781634191dd3c49097a615d2fc01b0a8aae31c448e55039a04dcbf36bba
curl -s -X POST \
  -H "Content-Type: application/json" \
  -d $'{
    "input": {
      "prompt": "Yoda",
      "num_outputs": 1,
      "guidance_scale": 7.5,
      "num_inference_steps": 50
    }
  }' \
  http://localhost:5000/predictions

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

Output

{
  "completed_at": "2022-09-21T09:44:49.500849Z",
  "created_at": "2022-09-21T09:44:19.304339Z",
  "data_removed": false,
  "error": null,
  "id": "n2vfz327yvgnpnrcbecukyqvh4",
  "input": {
    "prompt": "Yoda",
    "num_outputs": 1,
    "guidance_scale": 7.5,
    "num_inference_steps": "50"
  },
  "logs": "Using seed: 4358\n\n  0%|          | 0/51 [00:00<?, ?it/s]\n  2%|▏         | 1/51 [00:00<00:05,  9.21it/s]\n  4%|▍         | 2/51 [00:00<00:10,  4.84it/s]\n  6%|▌         | 3/51 [00:00<00:11,  4.22it/s]\n  8%|▊         | 4/51 [00:00<00:12,  3.89it/s]\n 10%|▉         | 5/51 [00:01<00:12,  3.79it/s]\n 12%|█▏        | 6/51 [00:01<00:12,  3.74it/s]\n 14%|█▎        | 7/51 [00:01<00:11,  3.70it/s]\n 16%|█▌        | 8/51 [00:02<00:11,  3.65it/s]\n 18%|█▊        | 9/51 [00:02<00:11,  3.64it/s]\n 20%|█▉        | 10/51 [00:02<00:11,  3.64it/s]\n 22%|██▏       | 11/51 [00:02<00:11,  3.62it/s]\n 24%|██▎       | 12/51 [00:03<00:10,  3.61it/s]\n 25%|██▌       | 13/51 [00:03<00:10,  3.61it/s]\n 27%|██▋       | 14/51 [00:03<00:10,  3.61it/s]\n 29%|██▉       | 15/51 [00:03<00:10,  3.59it/s]\n 31%|███▏      | 16/51 [00:04<00:09,  3.59it/s]\n 33%|███▎      | 17/51 [00:04<00:09,  3.60it/s]\n 35%|███▌      | 18/51 [00:04<00:09,  3.60it/s]\n 37%|███▋      | 19/51 [00:05<00:08,  3.59it/s]\n 39%|███▉      | 20/51 [00:05<00:08,  3.60it/s]\n 41%|████      | 21/51 [00:05<00:08,  3.60it/s]\n 43%|████▎     | 22/51 [00:05<00:08,  3.60it/s]\n 45%|████▌     | 23/51 [00:06<00:07,  3.61it/s]\n 47%|████▋     | 24/51 [00:06<00:07,  3.60it/s]\n 49%|████▉     | 25/51 [00:06<00:07,  3.60it/s]\n 51%|█████     | 26/51 [00:07<00:06,  3.60it/s]\n 53%|█████▎    | 27/51 [00:07<00:06,  3.60it/s]\n 55%|█████▍    | 28/51 [00:07<00:06,  3.59it/s]\n 57%|█████▋    | 29/51 [00:07<00:06,  3.59it/s]\n 59%|█████▉    | 30/51 [00:08<00:05,  3.60it/s]\n 61%|██████    | 31/51 [00:08<00:05,  3.60it/s]\n 63%|██████▎   | 32/51 [00:08<00:05,  3.60it/s]\n 65%|██████▍   | 33/51 [00:08<00:04,  3.60it/s]\n 67%|██████▋   | 34/51 [00:09<00:04,  3.61it/s]\n 69%|██████▊   | 35/51 [00:09<00:04,  3.60it/s]\n 71%|███████   | 36/51 [00:09<00:04,  3.61it/s]\n 73%|███████▎  | 37/51 [00:10<00:03,  3.60it/s]\n 75%|███████▍  | 38/51 [00:10<00:03,  3.60it/s]\n 76%|███████▋  | 39/51 [00:10<00:03,  3.61it/s]\n 78%|███████▊  | 40/51 [00:10<00:03,  3.61it/s]\n 80%|████████  | 41/51 [00:11<00:02,  3.60it/s]\n 82%|████████▏ | 42/51 [00:11<00:02,  3.60it/s]\n 84%|████████▍ | 43/51 [00:11<00:02,  3.61it/s]\n 86%|████████▋ | 44/51 [00:12<00:01,  3.62it/s]\n 88%|████████▊ | 45/51 [00:12<00:01,  3.61it/s]\n 90%|█████████ | 46/51 [00:12<00:01,  3.60it/s]\n 92%|█████████▏| 47/51 [00:12<00:01,  3.62it/s]\n 94%|█████████▍| 48/51 [00:13<00:00,  3.62it/s]\n 96%|█████████▌| 49/51 [00:13<00:00,  3.61it/s]\n 98%|█████████▊| 50/51 [00:13<00:00,  3.62it/s]\n100%|██████████| 51/51 [00:13<00:00,  3.63it/s]\n100%|██████████| 51/51 [00:13<00:00,  3.65it/s]",
  "metrics": {
    "predict_time": 15.236785,
    "total_time": 30.19651
  },
  "output": [
    "https://replicate.delivery/mgxm/4d12a241-fd84-4b0a-8321-80dd8c6ae784/out-0.png"
  ],
  "started_at": "2022-09-21T09:44:34.264064Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/n2vfz327yvgnpnrcbecukyqvh4",
    "cancel": "https://api.replicate.com/v1/predictions/n2vfz327yvgnpnrcbecukyqvh4/cancel"
  },
  "version": "3554d9e699e09693d3fa334a79c58be9a405dd021d3e11281256d53185868912"
}

Generated in

15.2 seconds

Tweak itReport

Using seed: 4358

  0%|          | 0/51 [00:00<?, ?it/s]
  2%|▏         | 1/51 [00:00<00:05,  9.21it/s]
  4%|▍         | 2/51 [00:00<00:10,  4.84it/s]
  6%|▌         | 3/51 [00:00<00:11,  4.22it/s]
  8%|▊         | 4/51 [00:00<00:12,  3.89it/s]
 10%|▉         | 5/51 [00:01<00:12,  3.79it/s]
 12%|█▏        | 6/51 [00:01<00:12,  3.74it/s]
 14%|█▎        | 7/51 [00:01<00:11,  3.70it/s]
 16%|█▌        | 8/51 [00:02<00:11,  3.65it/s]
 18%|█▊        | 9/51 [00:02<00:11,  3.64it/s]
 20%|█▉        | 10/51 [00:02<00:11,  3.64it/s]
 22%|██▏       | 11/51 [00:02<00:11,  3.62it/s]
 24%|██▎       | 12/51 [00:03<00:10,  3.61it/s]
 25%|██▌       | 13/51 [00:03<00:10,  3.61it/s]
 27%|██▋       | 14/51 [00:03<00:10,  3.61it/s]
 29%|██▉       | 15/51 [00:03<00:10,  3.59it/s]
 31%|███▏      | 16/51 [00:04<00:09,  3.59it/s]
 33%|███▎      | 17/51 [00:04<00:09,  3.60it/s]
 35%|███▌      | 18/51 [00:04<00:09,  3.60it/s]
 37%|███▋      | 19/51 [00:05<00:08,  3.59it/s]
 39%|███▉      | 20/51 [00:05<00:08,  3.60it/s]
 41%|████      | 21/51 [00:05<00:08,  3.60it/s]
 43%|████▎     | 22/51 [00:05<00:08,  3.60it/s]
 45%|████▌     | 23/51 [00:06<00:07,  3.61it/s]
 47%|████▋     | 24/51 [00:06<00:07,  3.60it/s]
 49%|████▉     | 25/51 [00:06<00:07,  3.60it/s]
 51%|█████     | 26/51 [00:07<00:06,  3.60it/s]
 53%|█████▎    | 27/51 [00:07<00:06,  3.60it/s]
 55%|█████▍    | 28/51 [00:07<00:06,  3.59it/s]
 57%|█████▋    | 29/51 [00:07<00:06,  3.59it/s]
 59%|█████▉    | 30/51 [00:08<00:05,  3.60it/s]
 61%|██████    | 31/51 [00:08<00:05,  3.60it/s]
 63%|██████▎   | 32/51 [00:08<00:05,  3.60it/s]
 65%|██████▍   | 33/51 [00:08<00:04,  3.60it/s]
 67%|██████▋   | 34/51 [00:09<00:04,  3.61it/s]
 69%|██████▊   | 35/51 [00:09<00:04,  3.60it/s]
 71%|███████   | 36/51 [00:09<00:04,  3.61it/s]
 73%|███████▎  | 37/51 [00:10<00:03,  3.60it/s]
 75%|███████▍  | 38/51 [00:10<00:03,  3.60it/s]
 76%|███████▋  | 39/51 [00:10<00:03,  3.61it/s]
 78%|███████▊  | 40/51 [00:10<00:03,  3.61it/s]
 80%|████████  | 41/51 [00:11<00:02,  3.60it/s]
 82%|████████▏ | 42/51 [00:11<00:02,  3.60it/s]
 84%|████████▍ | 43/51 [00:11<00:02,  3.61it/s]
 86%|████████▋ | 44/51 [00:12<00:01,  3.62it/s]
 88%|████████▊ | 45/51 [00:12<00:01,  3.61it/s]
 90%|█████████ | 46/51 [00:12<00:01,  3.60it/s]
 92%|█████████▏| 47/51 [00:12<00:01,  3.62it/s]
 94%|█████████▍| 48/51 [00:13<00:00,  3.62it/s]
 96%|█████████▌| 49/51 [00:13<00:00,  3.61it/s]
 98%|█████████▊| 50/51 [00:13<00:00,  3.62it/s]
100%|██████████| 51/51 [00:13<00:00,  3.63it/s]
100%|██████████| 51/51 [00:13<00:00,  3.65it/s]

This example was created by a different version, lambdal/text-to-pokemon:3554d9e6.

Examples

View more examples

Run time and cost

This model costs approximately $0.042 to run on Replicate, or 23 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 4 minutes. The predict time for this model varies significantly based on the inputs.

Readme

Stable Diffusion fine tuned on Pokémon by Lambda Labs.

Put in a text prompt and generate your own Pokémon character, no “prompt engineering” required!

If you want to find out how we made this model at Lambda Labs, read our blog post. Or if you want to train your own Stable Diffusion variants, see this example repo.

Girl with a pearl earring, Cute Obama creature, Donald Trump, Boris Johnson, Totoro, Hello Kitty

Model description

Trained on BLIP captioned Pokémon images using 2xA6000 GPUs on Lambda GPU Cloud for around 15,000 step (about 6 hours, at a cost of about $10).

Links

Trained by Justin Pinkney (@Buntworthy) at Lambda Labs.