Recommendation to Revisit the Diffuser Default LoRA Parameters
•
9
Minta K
alvdansen
AI & ML interests
Finetuning LoRA Models and ComfyUI Workflows.
If you want to use my LoRAs commercially, please reach out!
Articles
Organizations
alvdansen's activity
replied to
louisbrulenaudet's
post
2 days ago
Congrats :D
posted
an
update
3 days ago
Post
1996
Per popular request, I'm working on a beginning to end LoRA training workflow blog for a style.
It will focus on dataset curation through training on a pre-determined style to give a better insight on my process.
Curious what are some questions you might have that I can try to answer in it?
It will focus on dataset curation through training on a pre-determined style to give a better insight on my process.
Curious what are some questions you might have that I can try to answer in it?
posted
an
update
5 days ago
Post
2291
A few new styles added as SDXL LoRA:
Midsommar Cartoon
A playful cartoon style featuring bold colors and a retro aesthetic. Personal favorite at the moment.
alvdansen/midsommarcartoon
---
Wood Block XL
I've started training public domain styles to create some interesting datasets. In this case I found a group of images taken from really beautiful and colorful Japanese Blockprints.
alvdansen/wood-block-xl
--
Dimension W
For this model I did actually end up working on an SD 1.5 model as well as an SDXL. I prefer the SDXL version, and I am still looking for parameters I am really happy with for SD 1.5. That said, both have their merits. I trained this with the short film I am working on in mind.
alvdansen/dimension-w
alvdansen/dimension-w-sd15
Midsommar Cartoon
A playful cartoon style featuring bold colors and a retro aesthetic. Personal favorite at the moment.
alvdansen/midsommarcartoon
---
Wood Block XL
I've started training public domain styles to create some interesting datasets. In this case I found a group of images taken from really beautiful and colorful Japanese Blockprints.
alvdansen/wood-block-xl
--
Dimension W
For this model I did actually end up working on an SD 1.5 model as well as an SDXL. I prefer the SDXL version, and I am still looking for parameters I am really happy with for SD 1.5. That said, both have their merits. I trained this with the short film I am working on in mind.
alvdansen/dimension-w
alvdansen/dimension-w-sd15
replied to
their
post
8 days ago
I typically use Kohya, but I also test a lot of platform services for the right one because I am a creature of comfort :)
replied to
their
post
8 days ago
I need to double check the train_text_encoder_frac as I typically don't mess with that. For rank I'm usually at 32.
posted
an
update
8 days ago
Post
864
Hey all!
Here I take a somewhat strong stance and am petitioning to revisit the default training parameters on the Diffusers LoRA page.
In my opinion and after observing and testing may training pipelines shared by startups and resources, I have found that many of them exhibit the same types of issues. Upon discussing with some of these founders and creators, the common theme has been working backwards from the Diffusers LoRA page.
In this article, I explain why the defaults in the Diffuser LoRA code produce some positive results, which can be initially misleading, and a suggestion on how that could be improved.
https://huggingface.co/blog/alvdansen/revisit-diffusers-default-params
Here I take a somewhat strong stance and am petitioning to revisit the default training parameters on the Diffusers LoRA page.
In my opinion and after observing and testing may training pipelines shared by startups and resources, I have found that many of them exhibit the same types of issues. Upon discussing with some of these founders and creators, the common theme has been working backwards from the Diffusers LoRA page.
In this article, I explain why the defaults in the Diffuser LoRA code produce some positive results, which can be initially misleading, and a suggestion on how that could be improved.
https://huggingface.co/blog/alvdansen/revisit-diffusers-default-params
-
- 4 replies
replied to
their
post
9 days ago
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.
replied to
their
post
10 days ago
Here is fine also, I will check later
replied to
their
post
10 days ago
Are you on X? You can contact me there @araminta_k
replied to
their
post
10 days ago
I have added further observations here:
https://huggingface.co/blog/alvdansen/enhancing-lora-training-through-effective-captions
replied to
their
post
10 days ago
replied to
their
post
10 days ago
I will need to take a look at what the exact backend of face to all is. What is the result you’re getting ?
posted
an
update
11 days ago
Post
3164
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
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
-
- 13 replies
replied to
their
post
12 days ago
Thank you!
posted
an
update
13 days ago
Post
6693
I had a backlog of LoRA model weights for SDXL that I decided to prioritize this weekend and publish. I know many are using SD3 right now, however if you have the time to try them, I hope you enjoy them.
I intend to start writing more fully on the thought process behind my approach to curating and training style and subject finetuning, beginning this next week.
Thank you for reading this post! You can find the models on my page and I'll drop a few previews here.
I intend to start writing more fully on the thought process behind my approach to curating and training style and subject finetuning, beginning this next week.
Thank you for reading this post! You can find the models on my page and I'll drop a few previews here.
-
- 4 replies
thank you <3
🙌 Thank you so much for sharing! I will be sharing a training workflow in the coming week :D