Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/shyamsn97/stable-diffusion-server
Stable Diffusion colab server built with fastapi
https://github.com/shyamsn97/stable-diffusion-server
fastapi stable-diffusion
Last synced: 4 months ago
JSON representation
Stable Diffusion colab server built with fastapi
- Host: GitHub
- URL: https://github.com/shyamsn97/stable-diffusion-server
- Owner: shyamsn97
- License: mit
- Created: 2022-11-07T19:04:05.000Z (about 2 years ago)
- Default Branch: main
- Last Pushed: 2023-01-13T22:58:20.000Z (about 2 years ago)
- Last Synced: 2024-10-09T11:24:35.110Z (4 months ago)
- Topics: fastapi, stable-diffusion
- Language: Python
- Homepage:
- Size: 1.63 MB
- Stars: 7
- Watchers: 2
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Stable Diffusion Google Colab FastAPI Server
[](https://badge.fury.io/py/sd-server)
## Note: This is a pretty hacky server / client interface for generic Stable Diffusion pipelines using diffusers. It's not made for production use and hasn't really been optimized completely.
#### Installation
From pip:
```bash
pip install sd-server
```
From source:
```bash
git clone [email protected]:shyamsn97/stable-diffusion-server.git
cd stable-diffusion-server/
python setup.py install
```
### Stable Diffusion Server + Client Interface -- Simple Usage
Here we create a simple server hosting the standard `StableDiffusionPipeline` from the amazing package [`diffusers`](https://github.com/huggingface/diffusers). The client takes in `**kwargs` that should be the arguments passed into the `__call__` function from the pipeline. For instance, the `StableDiffusionPipeline` `__call__` method can be found: https://github.com/huggingface/diffusers/blob/main/src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion.py#L185. `host` and `port` can be specified in the `start` method for generic hosting.
```python
from sd_server import DiffusersServer, DiffusersClient
from diffusers import StableDiffusionPipeline
STABLE_DIFFUSION_PATH = "runwayml/stable-diffusion-v1-5" # this can be a local path
pipeline_kwargs = {
"revision":"fp16",
"torch_dtype":torch.float16
}
device = torch.device('cuda')
server = DiffusersServer.create(
pretrained_path = STABLE_DIFFUSION_PATH,
pipeline_cls = StableDiffusionPipeline,
pipeline_kwargs = pipeline_kwargs,
enable_attention_slicing = True,
device = device
)
url = server.start(host="127.0.0.1", port=8000) # url -- either remote or local
# on another host / terminal
client = DiffusersClient(url)
responses = client(prompt='a photo of an astronaut riding a horse on mars', num_images_per_prompt=4) # this should return a list of images
```
### Serving from Google Colab
Full google colab example: [](https://colab.research.google.com/drive/13B7EzuDTAsjd0hpUki_ZiqoQ4wgOyZUM?usp=sharing)
Using the server in Google Colab requires an [ngrok](https://ngrok.com/) account, and needs your ngrok auth token https://dashboard.ngrok.com/get-started/your-authtoken to be passed to the server. Note: Ngrok basically opens a tunnel from a dev machine to the cloud, which means its not really that secure, and should be used at your own risk. Read more here: https://stackoverflow.com/questions/36552950/is-ngrok-safe-to-use-or-can-it-be-compromised.
```python
NGROK_AUTH_KEY = "your auth key from https://dashboard.ngrok.com/get-started/your-authtoken"
... # imports from above
server = DiffusersServer.create(
pretrained_path = STABLE_DIFFUSION_PATH,
pipeline_cls = StableDiffusionPipeline,
pipeline_kwargs = pipeline_kwargs,
enable_attention_slicing = True,
device = device
)
url = server.start(ngrok_auth_token=NGROK_AUTH_KEY)
# on another host
client = DiffusersClient(url)
responses = client(prompt='a photo of an astronaut riding a horse on mars', num_images_per_prompt=4) # this should return a list of images
```