To achieve optimal results, consider incorporating both LoRA v.1.1 and v.1.0.
By blending these two models with the recommended weights of 0.45 for LoRA v.1.1 and 0.55 for LoRA v.1.0.
••••••••••••••••••••••••••••••••••••••••••••••••••••••
"Your support, no matter how small, can go a long way in keeping me fueled to create more valuable content. If you liked what I did and want to show your appreciation, why not buy me a kofi? It would mean a lot! Thank you for your kind consideration."
••••••••••••••••••••••••••••••••••••••••••••••••••••••
Look at the example image of LoRA v.1.1, which showcases the exceptional results that can be achieved by merging the two versions with the recommended weight.
These are simply suggestions and you are free to tailor them to your personal preferences.
*
And I apologize for any inconvenience, but I am unsure how to merge LoRA. Is there any way to merge them? Thank you.
••••••••••••••••••••••••••••••••••••••••••••••••••••••
To achieve the unique and captivating style of Hajime Sorayama, try using the trigger words "sorayama style".
••••••••••••••••••••••••••••••••••••••••••••••••••••••
I've made my very first LoRa. I'm still getting the hang of it, so any advice or recommendations you have would be incredibly valuable to me. And, of course, don't forget to share your artwork here! Thanks for your support and let's have some fun!
Description
FAQ
Comments (10)
This is looks amazing! - 10 stars !
Congratulations!
thanks for your recommendation mate!
@SEVUNX The only suggestions I have when making Lora is to use at least 50 to 75 HQ-HD large sized images for A.i. training. Images can be up to 2000 x 2000 pixels in large size. Train them up to 3000+ steps and only in batches up to 2. When making the Lora training select "enable buckets", so the program can use larger sized images. This way you will have alot of nice details in your A.i. rendered images. I do have plans to create a Hajime S. Lora in the future, because I am also a huge fan of his art and have about 8 big books with his art. So far, there are 4 Lora+TI already made (I have tested+used them all) and yours seems to be the best of the collection. Good job !
BTW - I really liked your previous 2 created Lora´s. I also like that type of art. Amazing results when using them !
@maxhitman Thank you for your helpful suggestions on creating Lora.
I'm glad to hear that you enjoy Hajime S. Lora's art and have plans to create your own in the future. Your compliments on my Lora creation and previous artworks are much appreciated. Thank you for your support!
Is there any way to specify which parts should be human and which parts should be left robotic?
I would like to share with you an image that was recently posted. As you can see in the posted image, it is about data generation. The community has shared a great prompt on the combination of human and robotic elements. I hope you enjoy it!
Thank you! Very happy to see it here.
Wait, you're not calling a lora in your prompt - so this is an embedding, not a lora, and should be listed as a textual inversion - goes in a different SD folder...?
hello, im using sd webui additional network by kohya ss
he probably gened these pics in a LoRA trainer. Those typically call the lora via inserting the path to the lora in the settings instead of calling it within the prompt. So when the images are created there's no need for the loras tag to be present. The trainer i use works this way too,
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.