|
import os |
|
import gradio as gr |
|
|
|
API_KEY=os.environ.get('HUGGING_FACE_HUB_TOKEN', None) |
|
|
|
article = """--- |
|
This space was created using [SD Space Creator](https://huggingface.co/spaces/anzorq/sd-space-creator).""" |
|
|
|
gr.Interface.load( |
|
name="models/darkstorm2150/Protogen_x3.4_Official_Release", |
|
title="""π€ No-code Gradio Demo Tutorial""", |
|
description="""Demo for <a href="https://huggingface.co/darkstorm2150/Protogen_x3.4_Official_Release">Protogen X3.4 Official Release</a> Stable Diffusion model. |
|
|
|
|
|
This space contains a Gradio demo that was created **without typing any code**! |
|
It enables you to generate photorealistic images based on a prompt. |
|
|
|
Here are the steps to take to create your own demo *in a few minutes* without any code: |
|
|
|
1. You can easily create a demo using [this HF space](https://huggingface.co/spaces/anzorq/sd-space-creator). You will have to provide a token with write access to continue as shown in the image below: |
|
|
|
<p align="center"> |
|
<img alt="gradio-demo-tutorial-1" src="https://huggingface.co/datasets/group2/tutorial-images/resolve/main/screenshots/1%20-%20gotToSpaceCreationl.PNG"> |
|
<em Enter a WRITE token.</em> |
|
</p> |
|
2. If your token is valid, a prompt will enable you to specify the URL or ID of the model on the HF Hub you would like to use in this demo. For example [darkstorm2150/Protogen_x3.4_Official_Release](https://huggingface.co/darkstorm2150/Protogen_x3.4_Official_Release), which is a diffusion model able to generate photorealistic images from a prompt. Click on "Load model". |
|
|
|
<p align="center"> |
|
<img alt="gradio-demo-tutorial-2" src="https://huggingface.co/datasets/group2/tutorial-images/resolve/main/screenshots/4-PasteTokenChooseModel.PNG"> |
|
<em Specify a model.</em> |
|
</p> |
|
3. Then, you can specify a few details about your demo, such as: |
|
- Name |
|
- Title |
|
- Description |
|
- Space type |
|
- Hardware (CPU? GPU?) |
|
Click on "Create the space". Your space will now be getting built. |
|
|
|
<p align="center"> |
|
<img alt="gradio-demo-tutorial-3" src="https://huggingface.co/datasets/group2/tutorial-images/resolve/main/screenshots/5-SpaceCreated.PNG"> |
|
<em Specify a model.</em> |
|
</p> |
|
4. Once your space is built, you can reach it at https://huggingface.co/spaces/SPACE_NAME. |
|
5. You can enter a prompt to generate an image with the space you just built! The demo embedded in this space was created that way without typing any code! |
|
<p align="center"> |
|
<img alt="gradio-demo-tutorial-5" src="https://huggingface.co/datasets/group2/tutorial-images/resolve/main/screenshots/6-GotYourNewSpace.PNG"> |
|
<em Specify a model.</em> |
|
</p> |
|
""", |
|
article=article, |
|
api_key=API_KEY, |
|
).queue(concurrency_count=20).launch() |
|
|