🚨 Training Flux LoRAs is the hottest trend in AI-enabled visual content. Check out this 4-min video where we guide you through training custom Flux models on #Scenario, making perfectly consistent characters with ease. The smoothest UI and workflow in the industry. YT link 📺 in the next post.

Also available via our YouTube channel

Thanks for creating this Emm. Incredible work!

Thanks! The credit goes to @rodrigon but the way. More videos coming!

What do you suggest is the best workflow when I don't have a dataset and generate the character with AI from scratch but still want it to be consistent? I've seen some people create character sheets (1 image depicting the character's face with various expressions) with Flux. Would it be best to generate the head and clothes separately? Or attempt a character sheet generation of the same character dressed from different angles and then separate it into several images for the training dataset?

@CamilleKleinman We have a video coming for this shortly. There are some workflows on the knowledge center too. You can create a character sheet in Flux, split it in parts to get 5-8 images, and train on them You can also use character reference in scenario to create variants

Can i download the trained lora for local inferences?

Yes

What if you want for interiors, best to create Lora for each usage and style? Ex bedroom, modern. How to do with dataset to create variety?

Up to you Some people create multiple Loras, each for a very specific style or purpose. Other make “general” Loras. Other merge Loras (like you can do on scenario btw) Smaller “focused” Loras are easier to train. You can always combine them later

Thank you for sharing