AI Engines is a Python library designed to seamlessly run a diverse range of AI models on demand – from text generation and image creation to TTS, text embeddings, image descriptors, and summarizers. It operates either through a user-friendly REST API or through a specialized non HTTP service that only loads specific engines.
My initial vision? To swiftly develop virtual personas I could interact with, not just through text but also vocally. I wanted them to be self-sufficient enough to handle their own "virtual space" – spanning across platforms like Twitter, Facebook, YouTube, emails, websites/blogs, and even interactions on Discord, Telegram, SMS, and more. I'll be providing more details soon, but in the meantime, give it a spin on revivya.com.
Enjoying the process of creating my #AI persona manager.
— Florian「 fy 」 (@un_geek) May 12, 2023
Current features:
🔀 Unlimited personas
✍️ Text generation (#GPT4, #ChatGPT, #GPTJ, #Llama)
🧠 Memory systems (#pgvector, #openai, #embeddings)
🎨 Image generation (#stablediffusion, #ControlNet)
🔖 Image… pic.twitter.com/RgFCHO2nDg