-
InfluxDB
Power Real-Time Data Analytics at Scale. Get real-time insights from all types of time series data with InfluxDB. Ingest, query, and analyze billions of data points in real-time with unbounded cardinality.
17 days ago I showed off a plugin I created for Krita that allowed you to run Stable Diffusion. It was only for linux and required you to install lots of python dependencies.
Now Stable Diffusion is being compiled with Pyinstaller and all its dependencies and I have added Windows support.
Now all you have to do is run the server and pass it a PID (this manual passing of a PID will be removed in the next release).
Here's how it works:
* I forked and cleaned up Stable Diffusion https://github.com/w4ffl35/stable-diffusion/tree/feature/win... that's what this version is using to compile it.
Additionally I released stablediffusiond 16 days ago
https://github.com/w4ffl35/stablediffusiond
it is a daemon for linux that runs Stable Diffusion and waits for requests along with a client daemon that handles requests.
currently it uses rabbitmq for queues. similar to sdrunner (which the krita plugin uses) it opens a socket and waits for connections.
sdrunner uses threads and SimpleQueue to achieve the same thing that stablediffusiond is doing with rabbitmq. they both use sockets.
stablediffusiond works very well.
if you're going to use something in prod, you should probably look into aws services, but stablediffusiond offers an alternative, and works really well locally.
Currently I am removing the PID requirement from sdrunner (which i will also move to a new repo).
this will make sdrunner a standalone local socket server which could be used by local devices to make stable diffusion requests that will generate images.
basically: it is a way for you to set up a machine in your house and have other machines (or itself) make requests to generate images. it has a queue and could accept as many clients as you like (only set up for 1 right now but this is changed with a single variable).