This model was trained as a test of Civitai Lora's trainer. It's my first Lora trained by Civitai trainer so it's not that good. to be completely honest I would say I failed. But I decided to post it anyway.Lora Weight: 0.6-0.7If your character is too sharp, you can use <a target="_blank" rel="ugc" href="https://civarchive.com/models/77203?modelVersionId=81979">Tangbohu Detail Tweaker LoRA </a>with weight of -1Remember this: The main problem with this version (which I won't abandon) is that Lora was trained on 45 Vitubers with different hairstyles and hair colors, so you need to prioritize the hair color in your prompt. Additionally, I strongly advise adding "multicolored hair" to your negatives. However, the biggest issue is the character's extremely fuzzy eyes. I highly recommend using Adetaler with an <a target="_blank" rel="ugc" href="https://civarchive.com/models/150925?modelVersionId=168820">eyes detector</a> and <a target="_blank" rel="ugc" href="https://civarchive.com/models/5529/eye-lora">Lora </a>for the eyes.Standard Checkpoint doesn't know what VTuber is (at list 1.5 don't), so I want to create a Checkpoint that would have an idea what it, is so training a VTuber Loras would be easier.

Files with SHA256: f28a1a2ae0d2e43fe7d6c64dd7af87c9226702727742ee69be5a660b8722e7b2