cswry/seesr – Run with an API on Replicate

cswry / seesr

SeeSR: Towards Semantics-Aware Real-World Image Super-Resolution

Cold

Public
68K runs
L40S
GitHub
Paper
License

Run with an API

Playground API Examples README Versions

Input

image

*file

Input image

user_prompt

string

Shift + Return to add a new line

Prompt to condition on

Default: ""

positive_prompt

string

Shift + Return to add a new line

Prompt to add

Default: "clean, high-resolution, 8k"

negative_prompt

string

Shift + Return to add a new line

Prompt to remove

Default: "dotted, noise, blur, lowres, smooth"

cfg_scale

number

(minimum: 0.1, maximum: 10)

Guidance scale, set value to >1 to use

Default: 5.5

num_inference_steps

integer

(minimum: 10, maximum: 100)

Number of inference steps

Default: 50

sample_times

integer

(minimum: 1, maximum: 10)

Number of samples to generate

Default: 1

latent_tiled_size

integer

(minimum: 128, maximum: 480)

Size of latent tiles

Default: 320

latent_tiled_overlap

integer

(minimum: 4, maximum: 16)

Overlap of latent tiles

Default: 4

scale_factor

integer

Scale factor

Default: 4

seed

integer

(minimum: 0, maximum: 2147483647)

Seed

Default: 231

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

npx create-replicate --model=cswry/seesr

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

const output = await replicate.run(
  "cswry/seesr:989cf3a66fd209363de347c3129d95d9fe639e44533ab47e07a6dfb3f250b6e3",
  {
    input: {
      seed: 231,
      image: "https://replicate.delivery/pbxt/KCSXe5GmB6MSQVpdD51dhe60nhlU91yA0MuDkXznmRt7QTXx/179.png",
      cfg_scale: 5.5,
      user_prompt: "",
      sample_times: 1,
      scale_factor: 4,
      negative_prompt: "dotted, noise, blur, lowres, smooth",
      positive_prompt: "clean, high-resolution, 8k",
      latent_tiled_size: 320,
      num_inference_steps: 50,
      latent_tiled_overlap: 4
    }
  }
);
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 cswry/seesr using Replicate’s API. Check out the model's schema for an overview of inputs and outputs.

output = replicate.run(
    "cswry/seesr:989cf3a66fd209363de347c3129d95d9fe639e44533ab47e07a6dfb3f250b6e3",
    input={
        "seed": 231,
        "image": "https://replicate.delivery/pbxt/KCSXe5GmB6MSQVpdD51dhe60nhlU91yA0MuDkXznmRt7QTXx/179.png",
        "cfg_scale": 5.5,
        "user_prompt": "",
        "sample_times": 1,
        "scale_factor": 4,
        "negative_prompt": "dotted, noise, blur, lowres, smooth",
        "positive_prompt": "clean, high-resolution, 8k",
        "latent_tiled_size": 320,
        "num_inference_steps": 50,
        "latent_tiled_overlap": 4
    }
)
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 cswry/seesr 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": "989cf3a66fd209363de347c3129d95d9fe639e44533ab47e07a6dfb3f250b6e3",
    "input": {
      "seed": 231,
      "image": "https://replicate.delivery/pbxt/KCSXe5GmB6MSQVpdD51dhe60nhlU91yA0MuDkXznmRt7QTXx/179.png",
      "cfg_scale": 5.5,
      "user_prompt": "",
      "sample_times": 1,
      "scale_factor": 4,
      "negative_prompt": "dotted, noise, blur, lowres, smooth",
      "positive_prompt": "clean, high-resolution, 8k",
      "latent_tiled_size": 320,
      "num_inference_steps": 50,
      "latent_tiled_overlap": 4
    }
  }' \
  https://api.replicate.com/v1/predictions

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

Output

{
  "completed_at": "2024-01-10T04:47:56.828664Z",
  "created_at": "2024-01-10T04:41:06.282807Z",
  "data_removed": false,
  "error": null,
  "id": "umyas7tbnchzkist6ys3n7xqsa",
  "input": {
    "seed": 231,
    "image": "https://replicate.delivery/pbxt/KCSXe5GmB6MSQVpdD51dhe60nhlU91yA0MuDkXznmRt7QTXx/179.png",
    "cfg_scale": 5.5,
    "user_prompt": "",
    "sample_times": 1,
    "scale_factor": 4,
    "negative_prompt": "dotted, noise, blur, lowres, smooth",
    "positive_prompt": "clean, high-resolution, 8k",
    "latent_tiled_size": 320,
    "num_inference_steps": 50,
    "latent_tiled_overlap": 4
  },
  "logs": "Seed set to 231\n[Tiled VAE]: the input size is tiny and unnecessary to tile.\n[Tiled Latent]: the input size is tiny and unnecessary to tile.\n  0%|          | 0/50 [00:00<?, ?it/s]\n  2%|▏         | 1/50 [00:00<00:09,  5.21it/s]\n  6%|▌         | 3/50 [00:00<00:04,  9.54it/s]\n 10%|█         | 5/50 [00:00<00:03, 11.27it/s]\n 14%|█▍        | 7/50 [00:00<00:03, 12.09it/s]\n 18%|█▊        | 9/50 [00:00<00:03, 12.49it/s]\n 22%|██▏       | 11/50 [00:00<00:03, 12.81it/s]\n 26%|██▌       | 13/50 [00:01<00:02, 13.05it/s]\n 30%|███       | 15/50 [00:01<00:02, 13.18it/s]\n 34%|███▍      | 17/50 [00:01<00:02, 13.29it/s]\n 38%|███▊      | 19/50 [00:01<00:02, 13.35it/s]\n 42%|████▏     | 21/50 [00:01<00:02, 13.38it/s]\n 46%|████▌     | 23/50 [00:01<00:02, 13.34it/s]\n 50%|█████     | 25/50 [00:01<00:01, 13.18it/s]\n 54%|█████▍    | 27/50 [00:02<00:01, 13.29it/s]\n 58%|█████▊    | 29/50 [00:02<00:01, 13.38it/s]\n 62%|██████▏   | 31/50 [00:02<00:01, 13.41it/s]\n 66%|██████▌   | 33/50 [00:02<00:01, 13.39it/s]\n 70%|███████   | 35/50 [00:02<00:01, 13.36it/s]\n 74%|███████▍  | 37/50 [00:02<00:00, 13.39it/s]\n 78%|███████▊  | 39/50 [00:03<00:00, 13.29it/s]\n 82%|████████▏ | 41/50 [00:03<00:00, 13.37it/s]\n 86%|████████▌ | 43/50 [00:03<00:00, 13.43it/s]\n 90%|█████████ | 45/50 [00:03<00:00, 13.48it/s]\n 94%|█████████▍| 47/50 [00:03<00:00, 13.48it/s]\n 98%|█████████▊| 49/50 [00:03<00:00, 13.47it/s]\n100%|██████████| 50/50 [00:03<00:00, 13.01it/s]\n[Tiled VAE]: the input size is tiny and unnecessary to tile.\n[Tiled VAE]: Done in 4.425s, max VRAM alloc 5669.202 MB",
  "metrics": {
    "predict_time": 6.151774,
    "total_time": 410.545857
  },
  "output": [
    "https://replicate.delivery/pbxt/Xx6RRm3aDkqtHNjOIvaxtxHRVjZwoF00KRRQNzmsUjMfmeKSA/0.png"
  ],
  "started_at": "2024-01-10T04:47:50.676890Z",
  "status": "succeeded",
  "urls": {
    "get": "https://api.replicate.com/v1/predictions/umyas7tbnchzkist6ys3n7xqsa",
    "cancel": "https://api.replicate.com/v1/predictions/umyas7tbnchzkist6ys3n7xqsa/cancel"
  },
  "version": "23b55f9ff1b4f0f254c3d8eb490d4b9d3e3d8961df1b7ea7bddf2641b411d6c9"
}

Generated in

6.2 seconds

Tweak itReport

This example was created by a different version, cswry/seesr:23b55f9f.

Examples

View more examples

Run time and cost

This model costs approximately $0.078 to run on Replicate, or 12 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 80 seconds. The predict time for this model varies significantly based on the inputs.

Readme

Cog Implementation of cswry/SeeSR

SeeSR: Towards Semantics-Aware Real-World Image Super-Resolution

1The Hong Kong Polytechnic University, 2OPPO Research Institute, 3ByteDance Inc.

⭐ If SeeSR is helpful to your images or projects, please help star this repo. Thanks!

Acknowledgments

This project is based on diffusers and BasicSR. Some codes are brought from PASD and RAM. Thanks for their awesome works. We also pay tribute to the pioneering work of StableSR.

Citations

If our code helps your research or work, please consider citing our paper. The following are BibTeX references:

@article{wu2023seesr,
  title={SeeSR: Towards Semantics-Aware Real-World Image Super-Resolution},
  author={Wu, Rongyuan and Yang, Tao and Sun, Lingchen and Zhang, Zhengqiang and Li, Shuai and Zhang, Lei},
  journal={arXiv preprint arXiv:2311.16518},
  year={2023}
}