nataili 0.3.1

Nataili: Multimodal AI Python Library

Nataili: Multimodal AI Python Library

!! NOTICE: Currently this package is looking for a lead developer !!

Our only ML Develeloper has dropped out and the reamining team does not have the required skills to add new features. If you see the value in this package, we urge you to contact us if you're interested in picking it up. You can reach us in discord or in this repository issues.

Nataili is a Python library that provides tools for building multimodal AI applications. With its modular design, Nataili makes it easy to use only the tools you need to build custom AI solutions.

Some of the technologies included in Nataili are:

• AITemplate: Fast diffusion
• BLIP: Image captioning
• CLIP: Interrogation (ranking, similarity), Embedding (text, image) with cache support
• CLIP+MLP Predictor
• CodeFormer: Post-processing and super-resolution for faces
• ESRGAN: Image super-resolution
• GFPGAN: Post-processing for faces
• LAION-CoCa: Image captioning
• Stable Diffusion: Image generation
• SD Upscaler: latent diffusion upscaler for Stable Diffusion's autoencoder
• Train: Training tools for Stable Diffusion and CLIP+MLP Predictor

Projects using Nataili

• AI Horde worker

Installation

pip install nataili

From source:

git clone https://github.com/db0/nataili
cd nataili
pip install -e .

Documentation

• Index
• Overview
• Model Manager
• Blip
• Clip
• LAION-CoCa