search

SamPatt / HaveWords.ai

P2P AI sharing using WebRTC
MIT License
5 stars 3 forks source link

Image generation and display #12

Open SamPatt opened 1 year ago

SamPatt commented 1 year ago

Hook up the ability to generate and display images.

DALLE-2 already works with the OpenAI API flow so that'll be first, but I would like to incorporate Stable Diffusion eventually, for people who run it locally.

Images can be called automatically during game flow, or randomly, or based on a set period of messages, or at the discretion of the host. Based on recent prompts, or responses, or a combination.

SamPatt commented 1 year ago

Hooked up the OpenAI DALLE-2 API, and it's working now.

Keeping open because I'd like to add other image options later.

SamPatt commented 1 year ago

DALLE-2 was removed because it sucked, but it should be added back in anticipation of DALLE-3, which should be accessible as soon as the API is offered (October?)