Huggingface model/space/data open source:

https://huggingface.co/models?pipeline_tag=image-to-image&sort=trending

Label classification:

https://editor.p5js.org/ml_4_cc/sketches/DbNwvgglU

Generate nft/train your own model from opensea

https://huggingface.co/spaces/huggan/huggingnft

How to train their own style of artistic model for artist?

using Dreambooth LoRA make small tweak on Stable Diffusion will need less images to get the vibe of the image style.

For now OpenArt has the best results on trainning model for artists. Just through in the pics, and train a style, the label or image classification is handled by other built-in LLMs.

https://www.recraft.ai/project/d80792f8-0813-4780-b549-959ed319ae0b

tried with recraft, training with only 5 pics, results are so sloppy:

https://appinventor.mit.edu/explore/ai-with-mit-app-inventor