charlesmccarthy/terminus-xl-gamma-v2 – Run with an API on Replicate

charlesmccarthy / terminus-xl-gamma-v2

Terminus XL Gamma is a new state-of-the-art latent diffusion model that uses zero-terminal SNR noise schedule and velocity prediction objective at training and inference time.

Cold

Public
279 runs
L40S

Run with an API

Playground API Examples README Versions

Input

prompt

string

Shift + Return to add a new line

joe biden eating a cheeseburger at mcdonaldsjoe biden eating a cheeseburger at mcdonalds

Input prompt

Default: "An astronaut riding a rainbow unicorn"

negative_prompt

string

Shift + Return to add a new line

Input Negative Prompt

Default: ""

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: 50)

Number of denoising steps

Default: 6

guidance_scale

number

(minimum: 1, maximum: 20)

Scale for classifier-free guidance

Default: 2

seed

integer

Random seed. Leave blank to randomize the seed

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

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](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=charlesmccarthy/terminus-xl-gamma-v2

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 charlesmccarthy/terminus-xl-gamma-v2 using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

const output = await replicate.run(
  "charlesmccarthy/terminus-xl-gamma-v2:991d0656366cbaec46d5457fe8994b1a09cca74f3f7590370d1a3669ad891189",
  {
    input: {
      width: 1024,
      height: 1024,
      prompt: "joe biden eating a cheeseburger at mcdonalds",
      scheduler: "DPMSolverMultistep",
      num_outputs: 1,
      guidance_scale: 6.5,
      apply_watermark: false,
      negative_prompt: "",
      num_inference_steps: 32
    }
  }
);
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 charlesmccarthy/terminus-xl-gamma-v2 using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "charlesmccarthy/terminus-xl-gamma-v2:991d0656366cbaec46d5457fe8994b1a09cca74f3f7590370d1a3669ad891189",
    input={
        "width": 1024,
        "height": 1024,
        "prompt": "joe biden eating a cheeseburger at mcdonalds",
        "scheduler": "DPMSolverMultistep",
        "num_outputs": 1,
        "guidance_scale": 6.5,
        "apply_watermark": False,
        "negative_prompt": "",
        "num_inference_steps": 32
    }
)
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 charlesmccarthy/terminus-xl-gamma-v2 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": "991d0656366cbaec46d5457fe8994b1a09cca74f3f7590370d1a3669ad891189",
    "input": {
      "width": 1024,
      "height": 1024,
      "prompt": "joe biden eating a cheeseburger at mcdonalds",
      "scheduler": "DPMSolverMultistep",
      "num_outputs": 1,
      "guidance_scale": 6.5,
      "apply_watermark": false,
      "negative_prompt": "",
      "num_inference_steps": 32
    }
  }' \
  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/charlesmccarthy/terminus-xl-gamma-v2@sha256:991d0656366cbaec46d5457fe8994b1a09cca74f3f7590370d1a3669ad891189 \
  -i 'width=1024' \
  -i 'height=1024' \
  -i 'prompt="joe biden eating a cheeseburger at mcdonalds"' \
  -i 'scheduler="DPMSolverMultistep"' \
  -i 'num_outputs=1' \
  -i 'guidance_scale=6.5' \
  -i 'apply_watermark=false' \
  -i 'negative_prompt=""' \
  -i 'num_inference_steps=32'

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/charlesmccarthy/terminus-xl-gamma-v2@sha256:991d0656366cbaec46d5457fe8994b1a09cca74f3f7590370d1a3669ad891189
curl -s -X POST \
  -H "Content-Type: application/json" \
  -d $'{
    "input": {
      "width": 1024,
      "height": 1024,
      "prompt": "joe biden eating a cheeseburger at mcdonalds",
      "scheduler": "DPMSolverMultistep",
      "num_outputs": 1,
      "guidance_scale": 6.5,
      "apply_watermark": false,
      "negative_prompt": "",
      "num_inference_steps": 32
    }
  }' \
  http://localhost:5000/predictions

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

Output

{
  "completed_at": "2023-12-26T05:21:12.504487Z",
  "created_at": "2023-12-26T05:21:04.247192Z",
  "data_removed": false,
  "error": null,
  "id": "uw5lrwtbwdi527is4mpwjgly4q",
  "input": {
    "width": 1024,
    "height": 1024,
    "prompt": "joe biden eating a cheeseburger at mcdonalds",
    "scheduler": "DPMSolverMultistep",
    "num_outputs": 1,
    "guidance_scale": 6.5,
    "apply_watermark": false,
    "negative_prompt": "",
    "num_inference_steps": 32
  },
  "logs": "Using seed: 10943700\n  0%|          | 0/32 [00:00<?, ?it/s]\n  3%|▎         | 1/32 [00:00<00:06,  5.02it/s]\n  9%|▉         | 3/32 [00:00<00:04,  6.21it/s]\n 12%|█▎        | 4/32 [00:00<00:04,  5.75it/s]\n 16%|█▌        | 5/32 [00:00<00:04,  5.49it/s]\n 19%|█▉        | 6/32 [00:01<00:04,  5.32it/s]\n 22%|██▏       | 7/32 [00:01<00:04,  5.21it/s]\n 25%|██▌       | 8/32 [00:01<00:04,  5.15it/s]\n 28%|██▊       | 9/32 [00:01<00:04,  5.11it/s]\n 31%|███▏      | 10/32 [00:01<00:04,  5.08it/s]\n 34%|███▍      | 11/32 [00:02<00:04,  5.06it/s]\n 38%|███▊      | 12/32 [00:02<00:03,  5.04it/s]\n 41%|████      | 13/32 [00:02<00:03,  5.03it/s]\n 44%|████▍     | 14/32 [00:02<00:03,  5.02it/s]\n 47%|████▋     | 15/32 [00:02<00:03,  5.02it/s]\n 50%|█████     | 16/32 [00:03<00:03,  5.01it/s]\n 53%|█████▎    | 17/32 [00:03<00:02,  5.01it/s]\n 56%|█████▋    | 18/32 [00:03<00:02,  5.01it/s]\n 59%|█████▉    | 19/32 [00:03<00:02,  5.00it/s]\n 62%|██████▎   | 20/32 [00:03<00:02,  5.00it/s]\n 66%|██████▌   | 21/32 [00:04<00:02,  5.00it/s]\n 69%|██████▉   | 22/32 [00:04<00:01,  5.00it/s]\n 72%|███████▏  | 23/32 [00:04<00:01,  4.99it/s]\n 75%|███████▌  | 24/32 [00:04<00:01,  4.99it/s]\n 78%|███████▊  | 25/32 [00:04<00:01,  5.00it/s]\n 81%|████████▏ | 26/32 [00:05<00:01,  4.99it/s]\n 84%|████████▍ | 27/32 [00:05<00:01,  4.99it/s]\n 88%|████████▊ | 28/32 [00:05<00:00,  4.99it/s]\n 91%|█████████ | 29/32 [00:05<00:00,  4.99it/s]\n 94%|█████████▍| 30/32 [00:05<00:00,  4.99it/s]\n 97%|█████████▋| 31/32 [00:06<00:00,  4.99it/s]\n100%|██████████| 32/32 [00:06<00:00,  4.99it/s]\n100%|██████████| 32/32 [00:06<00:00,  5.08it/s]",
  "metrics": {
    "predict_time": 8.221049,
    "total_time": 8.257295
  },
  "output": [
    "https://replicate.delivery/pbxt/WqQWvzc3p1pqNJfhVc0Tz5xwTVulnxEoTylie1yKCZWHTBGSA/out-0.png"
  ],
  "started_at": "2023-12-26T05:21:04.283438Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/uw5lrwtbwdi527is4mpwjgly4q",
    "cancel": "https://api.replicate.com/v1/predictions/uw5lrwtbwdi527is4mpwjgly4q/cancel"
  },
  "version": "991d0656366cbaec46d5457fe8994b1a09cca74f3f7590370d1a3669ad891189"
}

Generated in

8.2 seconds

Tweak itReport

Examples

View more examples

Run time and cost

This model costs approximately $0.0081 to run on Replicate, or 123 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 9 seconds. The predict time for this model varies significantly based on the inputs.

Readme

Terminus XL Gamma is a new state-of-the-art latent diffusion model that uses zero-terminal SNR noise schedule and velocity prediction objective at training and inference time.

Terminus is based on a similar architecture to SDXL, and has the same layout. It has been trained on fewer steps with very high quality data captions via COCO and Midjourney.

This model will not be capable of as many concepts as SDXL, and some subjects will simply look very bad.

The objective of this model was to use v-prediction and min-SNR gamma loss to efficiently train a full zero-terminal SNR model on a single A100-80G.

Fine-tuned from: ptx0/terminus-xl-gamma-v1 Developed by: pseudoterminal X (@bghira) Funded by: pseudoterminal X (@bghira) Model type: Latent Diffusion License: openrail++ Architecture: SDXL