README.md

---
library_name: diffusers
tags:
- stable-diffusion-3.5
- diffusers
- text-to-image
- endpoints-template
inference: false
---

> [!IMPORTANT]
> This repo duplicates the [Stable Diffusion 3.5 Large](https://huggingface.co/stabilityai/stable-diffusion-3.5-large) weights for demonstration purposes and it doesn't own any credits to the model. So, please be mindful of that and respect the original license of the model.

```py
from handler import EndpointHandler

my_handler = EndpointHandler()
payload = {"inputs": {"prompt": "a dog waiting for its companion to come."}}

# test the handler
image = my_handler(payload)
image.save("image.png")
```

![](./image.png)

We can use this repo to deploy SD3.5 Large on [Inference Endpoints](https://huggingface.co/docs/inference-endpoints) as well.

```py
import json
import requests
import base64
from PIL import Image
from io import BytesIO

ENDPOINT_URL = ""
HF_TOKEN = ""


def decode_base64_image(image_string):
    base64_image = base64.b64decode(image_string)
    buffer = BytesIO(base64_image)
    return Image.open(buffer).save("image.png")


def predict(prompt: str = "a dog waiting for its companion to come."):
    payload = {"inputs": {"prompt": prompt}}
    response = requests.post(ENDPOINT_URL, headers={"Authorization": f"Bearer {HF_TOKEN}"}, json=payload)
    resp = response.json()
    return decode_base64_image(resp)


prediction = predict(prompt="the first animal on the mars")
```

This is how the final image would look like:

![](./ie_image.png)