All my models are available here
https://www.buymeacoffee.com/razzzp/posts
Disclaimer: This artificial intelligence model ("the Model") is developed solely for artistic expression and technological demonstration, not for creating or distributing illicit or inappropriate content. The creator expressly disapproves of such use. By downloading or using the Model, you ("the User") agree not to misuse it in any manner contrary to these intentions and accept all responsibility for compliance with these terms.
Description
Initial Release
FAQ
Comments (16)
Razzz were I can buy you a coffee ?
this is really good, but I do have a question
since it is 9200 steps, how long did it take to train? (I assume it was on 4090?)
Does it work also with the SDXL refiner? I trained a lora on the sdxl base model myself and while it works quite well with a workflow just with the base model, once I create a workflow with added refiner the output deviates massively from the character.
razzz, this lora is excellent. Could you please share your training command. I can't figure out how you got 9200 steps with 39 images and 6 epochs
Great LoRA! Can you help with some questions, please?
1. Can you share your LoRA training config or accelerate command?
2. What regularization images do you use?
3. Why do you use train batch size of 1? Having 4090 you could increase this value and significantly decrease training time.
This lora is amazing! What's the size of the dataset you use?
"before they mandated celebrities to wear a hijab."
LOL. Hope you're wrong and this place doesn't go that route.
You are as fast as lightning! As soon as Barbie was released, you managed to train Lora
Love it. Can you make a SDXL Lora for Dasha Taran? She's very popular, and Aitreprenuer has a complex guide with the best strategies on how to make it, so idk if it can help.
Hey bud. Awesome LORA! I am struggling to create a lora like this. could you tell me... how many base images did you use? how many face? how many body? dim, alpha, repeats, epochs, batch. all that. can you tell us your formula? thanks.!
Awesome Lora!
Is it better to train these celebrity Loras on the base SDXL model, or on the custom checkpoint I plan to use them on?
@razz you've been making some cool 2.0s on some models, is this one still up to par with your new methods? :)
Also, deborah ann woll next? :D
hard to generate full body, or I am doing something wrong :(.
Hoping for 2.0 on this one and Gal Gadot :D You're the best @Razz :D
Sidenote, did you try using your datasets to make loras for PonyDiffusion? Because I noticed pony just doesn't work with the sdxl 1.0, and nobody, ( i mean nobody) has been able to tell me if pony can actually capture the liekness of a REAL person if you try to train it for it, so I just don't know
much rather have the superior young jaime pressly
Details
Files
margot_lora_sdxl_v1-000006.safetensors
Mirrors
margot_lora_sdxl_v1-000006.safetensors
93417_margot_lora_sdxl_v1-000006.safetensors
margot_lora_sdxl_v1-000006.safetensors
margot_lora_sdxl_v1-000006.safetensors
margot_lora_sdxl_v1-000006.safetensors
margot_lora_sdxl_v1-000006.safetensors
margot_lora_sdxl_v1-000006.safetensors
margot_lora_sdxl_v1-000006.safetensors
margot_lora_sdxl_v1-000006.safetensors
margot_lora_sdxl_v1-000006.safetensors
margot_lora_sdxl_v1-000006.safetensors
Margot_Robbie.safetensors
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.










