This lora is to generate auroral background. Each version is independent.
This basic model is realism, and dataset is no-human, it is better to use this lora in background generation. Each version have different dataset, you can chose one you like.
You can use Hires.fix to upscale.
V.1 & Lighting auroral: Dataset:512*512
Others has 960*960 dataset for train
Trigger words: Auroral, ice, glaciers.Lightning, electronic
Space Hole trigger word: space, galaxy
suggested prompt key words: light, night, mushroom, wave, lake,sea,wind
Please give me 5 stars and like π, which is important to me.
πyour reviews motivate me to updateοΌyour suggestion for updates are also helpfulοΌ
π π π π ππ π π π ππ π π π ππ π π π ππ π π π ππ π π π ππ π π π π
Hope any feedback from you, you can link me onπΈ discord, huggingface, github.
πΈAll train data collection are human-manual, and training cloud environment costs me much. hope your donation to support my new model. π€https://ko-fi.com/cecily34
If you have any questions related to the model or its samples, how to get started or just want to share your creations, feel free to join my discord:Β https://discord.gg/GgZFF3qPEB
That is my pleasure.
π ( β’Μ₯Μ α¦· β’Μ ) Loves U.
Description
No trigger words
Dataset:960*960
FAQ
Comments (7)
This seems super overtuned, even at 0.1 it's completely taking over prompts.
The dataset of this lora is just scene, if you want to focus and highlight characters in the center of the image, it cannot perform greatly. because the dataset image is very focusing on scene, not character.
@Cecily_ccΒ still clearly its overtuned
@Klara88Β which one
@Cecily_ccΒ not tested all yet, but space hole is indeed very overtuned taking over everything else even at 0.1. Will try 0.01 to see.
@yoda5477Β thank you
@Cecily_ccΒ You removed the space hole one ?