RS-CVN5 is an anime-style LoRA trained on Anima Base v1.0, with soft shading, vivid colors, detailed eyes, and expressive character portraits.
Usage
Base model: Anima Base v1.0
Trigger word: rs-cvn5
Tested LoRA strength: 1.0 (adjust to taste)
Sample settings
ComfyUI · 1024 × 1024 · 24 steps · CFG 4 · Euler sampler · simple scheduler
All showcase images were generated with this LoRA at strength 1.0. The gallery focuses on adult female characters across everyday, fantasy, and fashion outfits.
Training
Final checkpoint: 11,710 steps / 10 epochs
LoRA rank: 32 · alpha: 12
Installation in ComfyUI
Place RS-CVN5.safetensors in ComfyUI/models/loras/, load it with an Anima-compatible LoRA loader, and add rs-cvn5 to your positive prompt.
This is the first release. Results depend on the prompt and seed; hands and fine details may still need adjustment.
Description
Initial release, trained on Anima Base v1.0.
Final checkpoint: 11,710 steps / 10 epochs; rank 32, alpha 12.
Trigger: rs-cvn5. Tested strength: 1.0.
Showcase settings: ComfyUI, 1024×1024, Euler / simple, 24 steps, CFG 4.