Divided by Zer0
Divided by Zer0
We already support setting generations as shared. We should report that to the worker. This way worker owners who want to, can use those generations for other purposes (analysis, or...
Currently all our styles are based on a json file on github. It might be worth considering integrating into the AI Horde. * Someone could create one or more style...
trusted/patreon users should be allowed to bypass the 2/sec ratelimit
It might be cool to connect AI Horde integrations to AI Horde accounts. Way this could work: 1 new table with integrations. `ID(int) | user_id(int, foreign key) | name(string) |...
Slow worker being one that is at < 0.3 MPS/s Ideas: - [x] Add a switch for people to request to exclude slow workers from their generations. Provide a kudos...
This allows a holistic adjustment based on all parameters
I've put a lot of effort in trying to prevent a race condition between workers picking up the same request with the same seed, and therefore returning two identical images....
https://petals.ml/ Set up a petal client on the horde which can talk to the bloom. Set up a rest API endpoint on the horde for people to request gens on...
Might be useful for helping workers filter https://beta.openai.com/docs/api-reference/fine-tunes/cancel
Starting this so I can have somewhere to post links https://pypi.org/project/sewar/