same as the other malay model in the collection.

however, the whole point of image generations is simply to play around.

use lora tag manager in comfyui to pick out all the various tags, but do note most of the words does not really help as they detract from the actual item you'd like to see appear in your image.

have fun testing out all the possible variations~!

https://civitai.com/posts/12110717

https://civitai.com/posts/12109990

learn by remixing the images from above civitai post!

my training settings i've changed the following which results in a larger file.

more learned information = more data stored

[additional_network_arguments]

unet_lr = 0.0005

text_encoder_lr = 5e-5

network_dim = 128

network_alpha = 16

network_module = "networks.lora"