anotherjesse/sdv2-preview – Run with an API on Replicate

anotherjesse / sdv2-preview

Stable Diffusion 2.0 Preview

Cold

Public
28.4K runs
A100 (80GB)
GitHub
License

Run with an API

Playground API Examples README Versions

Input

prompt

string

Shift + Return to add a new line

an concept art of the cat king, pale hair, one eye, intricate details, detailed face, detailed armour, artstation, epic pose, ambient light, by eiichiro oda an concept art of the cat king, pale hair, one eye, intricate details, detailed face, detailed armour, artstation, epic pose, ambient light, by eiichiro oda 

Input prompt

Default: ""

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

num_outputs

integer

(minimum: 1, maximum: 10)

Number of images to output. If the NSFW filter is triggered, you may get fewer outputs than this.

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

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=anotherjesse/sdv2-preview

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 anotherjesse/sdv2-preview using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "anotherjesse/sdv2-preview:9d608f854eed9805b26cdc3514a625d32f30f2c50075eb673f0c6541635c8c5e",
  {
    input: {
      width: 512,
      height: 512,
      prompt: "an concept art of the cat king, pale hair, one eye, intricate details, detailed face, detailed armour, artstation, epic pose, ambient light, by eiichiro oda ",
      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 anotherjesse/sdv2-preview using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "anotherjesse/sdv2-preview:9d608f854eed9805b26cdc3514a625d32f30f2c50075eb673f0c6541635c8c5e",
    input={
        "width": 512,
        "height": 512,
        "prompt": "an concept art of the cat king, pale hair, one eye, intricate details, detailed face, detailed armour, artstation, epic pose, ambient light, by eiichiro oda ",
        "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 anotherjesse/sdv2-preview 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": "9d608f854eed9805b26cdc3514a625d32f30f2c50075eb673f0c6541635c8c5e",
    "input": {
      "width": 512,
      "height": 512,
      "prompt": "an concept art of the cat king, pale hair, one eye, intricate details, detailed face, detailed armour, artstation, epic pose, ambient light, by eiichiro oda ",
      "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/anotherjesse/sdv2-preview@sha256:9d608f854eed9805b26cdc3514a625d32f30f2c50075eb673f0c6541635c8c5e \
  -i 'width=512' \
  -i 'height=512' \
  -i 'prompt="an concept art of the cat king, pale hair, one eye, intricate details, detailed face, detailed armour, artstation, epic pose, ambient light, by eiichiro oda "' \
  -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/anotherjesse/sdv2-preview@sha256:9d608f854eed9805b26cdc3514a625d32f30f2c50075eb673f0c6541635c8c5e
curl -s -X POST \
  -H "Content-Type: application/json" \
  -d $'{
    "input": {
      "width": 512,
      "height": 512,
      "prompt": "an concept art of the cat king, pale hair, one eye, intricate details, detailed face, detailed armour, artstation, epic pose, ambient light, by eiichiro oda ",
      "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": "2022-11-24T06:03:40.523373Z",
  "created_at": "2022-11-24T06:03:35.999117Z",
  "data_removed": false,
  "error": null,
  "id": "rio3gcv6indwrk5hk65cu7op5i",
  "input": {
    "width": "512",
    "height": 512,
    "prompt": "an concept art of the cat king, pale hair, one eye, intricate details, detailed face, detailed armour, artstation, epic pose, ambient light, by eiichiro oda ",
    "num_outputs": "1",
    "guidance_scale": 7.5,
    "num_inference_steps": 50
  },
  "logs": "Using seed: 32970\nGlobal seed set to 32970\nData shape for DDIM sampling is (1, 4, 64, 64), eta 0.0\nRunning DDIM Sampling with 50 timesteps\nDDIM Sampler:   0%|          | 0/50 [00:00<?, ?it/s]\nDDIM Sampler:   4%|▍         | 2/50 [00:00<00:03, 13.25it/s]\nDDIM Sampler:   8%|▊         | 4/50 [00:00<00:03, 14.01it/s]\nDDIM Sampler:  12%|█▏        | 6/50 [00:00<00:03, 14.41it/s]\nDDIM Sampler:  16%|█▌        | 8/50 [00:00<00:03, 13.96it/s]\nDDIM Sampler:  20%|██        | 10/50 [00:00<00:02, 14.16it/s]\nDDIM Sampler:  24%|██▍       | 12/50 [00:00<00:02, 14.23it/s]\nDDIM Sampler:  28%|██▊       | 14/50 [00:00<00:02, 14.38it/s]\nDDIM Sampler:  32%|███▏      | 16/50 [00:01<00:02, 14.33it/s]\nDDIM Sampler:  36%|███▌      | 18/50 [00:01<00:02, 14.19it/s]\nDDIM Sampler:  40%|████      | 20/50 [00:01<00:02, 14.13it/s]\nDDIM Sampler:  44%|████▍     | 22/50 [00:01<00:01, 14.23it/s]\nDDIM Sampler:  48%|████▊     | 24/50 [00:01<00:01, 14.33it/s]\nDDIM Sampler:  52%|█████▏    | 26/50 [00:01<00:01, 14.39it/s]\nDDIM Sampler:  56%|█████▌    | 28/50 [00:01<00:01, 14.27it/s]\nDDIM Sampler:  60%|██████    | 30/50 [00:02<00:01, 14.35it/s]\nDDIM Sampler:  64%|██████▍   | 32/50 [00:02<00:01, 14.44it/s]\nDDIM Sampler:  68%|██████▊   | 34/50 [00:02<00:01, 14.39it/s]\nDDIM Sampler:  72%|███████▏  | 36/50 [00:02<00:00, 14.56it/s]\nDDIM Sampler:  76%|███████▌  | 38/50 [00:02<00:00, 14.69it/s]\nDDIM Sampler:  80%|████████  | 40/50 [00:02<00:00, 14.41it/s]\nDDIM Sampler:  84%|████████▍ | 42/50 [00:02<00:00, 14.49it/s]\nDDIM Sampler:  88%|████████▊ | 44/50 [00:03<00:00, 14.51it/s]\nDDIM Sampler:  92%|█████████▏| 46/50 [00:03<00:00, 14.59it/s]\nDDIM Sampler:  96%|█████████▌| 48/50 [00:03<00:00, 14.53it/s]\nDDIM Sampler: 100%|██████████| 50/50 [00:03<00:00, 14.48it/s]\nDDIM Sampler: 100%|██████████| 50/50 [00:03<00:00, 14.36it/s]",
  "metrics": {
    "predict_time": 4.488776,
    "total_time": 4.524256
  },
  "output": [
    "https://replicate.delivery/pbxt/3JlNy0OCE6JZFN2huaoJFjfGO9tDRHRDQs6QIWeRoEP7sLDQA/out-0.png"
  ],
  "started_at": "2022-11-24T06:03:36.034597Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/rio3gcv6indwrk5hk65cu7op5i",
    "cancel": "https://api.replicate.com/v1/predictions/rio3gcv6indwrk5hk65cu7op5i/cancel"
  },
  "version": "9d608f854eed9805b26cdc3514a625d32f30f2c50075eb673f0c6541635c8c5e"
}

Generated in

4.5 seconds

Tweak itReport

Using seed: 32970
Global seed set to 32970
Data shape for DDIM sampling is (1, 4, 64, 64), eta 0.0
Running DDIM Sampling with 50 timesteps
DDIM Sampler:   0%|          | 0/50 [00:00<?, ?it/s]
DDIM Sampler:   4%|▍         | 2/50 [00:00<00:03, 13.25it/s]
DDIM Sampler:   8%|▊         | 4/50 [00:00<00:03, 14.01it/s]
DDIM Sampler:  12%|█▏        | 6/50 [00:00<00:03, 14.41it/s]
DDIM Sampler:  16%|█▌        | 8/50 [00:00<00:03, 13.96it/s]
DDIM Sampler:  20%|██        | 10/50 [00:00<00:02, 14.16it/s]
DDIM Sampler:  24%|██▍       | 12/50 [00:00<00:02, 14.23it/s]
DDIM Sampler:  28%|██▊       | 14/50 [00:00<00:02, 14.38it/s]
DDIM Sampler:  32%|███▏      | 16/50 [00:01<00:02, 14.33it/s]
DDIM Sampler:  36%|███▌      | 18/50 [00:01<00:02, 14.19it/s]
DDIM Sampler:  40%|████      | 20/50 [00:01<00:02, 14.13it/s]
DDIM Sampler:  44%|████▍     | 22/50 [00:01<00:01, 14.23it/s]
DDIM Sampler:  48%|████▊     | 24/50 [00:01<00:01, 14.33it/s]
DDIM Sampler:  52%|█████▏    | 26/50 [00:01<00:01, 14.39it/s]
DDIM Sampler:  56%|█████▌    | 28/50 [00:01<00:01, 14.27it/s]
DDIM Sampler:  60%|██████    | 30/50 [00:02<00:01, 14.35it/s]
DDIM Sampler:  64%|██████▍   | 32/50 [00:02<00:01, 14.44it/s]
DDIM Sampler:  68%|██████▊   | 34/50 [00:02<00:01, 14.39it/s]
DDIM Sampler:  72%|███████▏  | 36/50 [00:02<00:00, 14.56it/s]
DDIM Sampler:  76%|███████▌  | 38/50 [00:02<00:00, 14.69it/s]
DDIM Sampler:  80%|████████  | 40/50 [00:02<00:00, 14.41it/s]
DDIM Sampler:  84%|████████▍ | 42/50 [00:02<00:00, 14.49it/s]
DDIM Sampler:  88%|████████▊ | 44/50 [00:03<00:00, 14.51it/s]
DDIM Sampler:  92%|█████████▏| 46/50 [00:03<00:00, 14.59it/s]
DDIM Sampler:  96%|█████████▌| 48/50 [00:03<00:00, 14.53it/s]
DDIM Sampler: 100%|██████████| 50/50 [00:03<00:00, 14.48it/s]
DDIM Sampler: 100%|██████████| 50/50 [00:03<00:00, 14.36it/s]

Examples

View more examples

Run time and cost

This model costs approximately $0.044 to run on Replicate, or 22 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 A100 (80GB) GPU hardware. Predictions typically complete within 32 seconds. The predict time for this model varies significantly based on the inputs.

Readme

Model description

This is a first pass at converting the new Stable Diffusion 2.0 released November 23rd to work on replicate.

This is using the Base weights.

Caveats and recommendations

this is a quick hack to see how well the code posted by StableAI works on replicate. This will probably be replaced by an official version by the Replicate team.