Hugging Face's logo

Join the conversation

Join the community of Machine Learners and AI enthusiasts.

Sign Up
Back to feed
alvdansen 
posted an update Jun 18
Post
3247
Hey All!

I've been asked a lot of share more on how I train LoRAs. The truth is I don't think my advice is very helpful without also including more contextual, theoretical commentary on how I **think** about training LoRAs for SDXL and other models.

I wrote a first article here about it - let me know what you think.

https://huggingface.co/blog/alvdansen/thoughts-on-lora-training-1

Edit: Also people kept asking where to start so I made a list of possible resources:
https://huggingface.co/blog/alvdansen/thoughts-on-lora-training-pt-2-training-services
Viktor1233

Hello! I found your lora in face to all, one of the best. But I can't adapt your other loras for Face to All, they don't work properly. I'm new and maybe I don't understand what to do.

·
alvdansen

I will need to take a look at what the exact backend of face to all is. What is the result you’re getting ?

alvdansen

I have added further observations here:

https://huggingface.co/blog/alvdansen/enhancing-lora-training-through-effective-captions

Viewegger

Just to be sure, since you don't mention any hyperparameters in your post does that mean that you are always using the default ones?

·
alvdansen

No - I change them however it’s very case by case. I am trying to emphasize elements other than hyperparameters, because in my experience these concepts apply to a variety of hyperparameters.

Fihade

What is your favorite and most used LoRA Trainer tool?

In this post