
I was a big fan of Lion-O, I mean, who wasn't? I wanted to pay homage to one of my favourite childhood heroes so training this LoRa was a cool experience, I had a lot of fun. This model is very flexible as shown in the examples so feel free to try it and share your creations!
Context:
Lion-O, the valiant Lord of the ThunderCats, roared onto screens in 1985, wielding the legendary Sword of Omens against the evil Mumm-Ra. With his loyal ThunderCat companions and trusty feline sidekick Snarf, Lion-O defended Third Earth, embodying courage, justice, and the power of the Eye of Thundera. This iconic cartoon hero, with his flowing red mane and lightning bolt battle cry, captivated audiences and left an indelible mark on pop culture history.
Usage:
Activation Words: Lion-O man
Recommeded settings:
LoRa Weight: 0.7. Please use this weight if you want to ensure te best likeness and output.
Clip Skip: 1
Sampler: All the DPM++ SDE work great. I personally prefer DPM++ 3M SDE Karras
Resolution: at least 1024 x 1024 px
Use ADetailer if you want but try to not go over 0.4 denoising strenght, 0.3 recommended.
If your hardware allows it, give it a 1.5 hires fix with a denoising of 0.2 - 0.3 and you will get much more details.
Upscaler: 2x-AnimeClassics-UltraLite if you go for the original ‘85 style, else 4xNMKD-Superscale-SP_178000_G or 4xNMKD-Superscale-_150000_G
I made the sample images using Forge which is similar to Automatic 1111 but it works the same in ComfyUI
描述:
v1
训练词语: Lion-O man
名称: Lion-O-v1-000003.safetensors
大小 (KB): 445810
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success