animagine-xl-4.0-opt-pvc-style-lora Open Source Model - Free Generation of PVC Figure Style Anime Character Images

Animagine Xl 4.0 Opt Pvc Style Lora

Developed by p1atdev

A LoRA model optimized based on AnimagineXL 4.0, specifically designed for generating anime character images in PVC figure style

Image Generation Open Source License:Apache-2.0 #PVC Figure Style #Anime Character Generation #LoRA Fine-tuning

Downloads 63

Release Time : 4/10/2025

Model Overview

This model is an optimized version of AnimagineXL 4.0, fine-tuned using LoRA technology to focus on generating high-quality PVC figure-style anime character images. Supports multiple trigger words to produce different figure effects.

Model Features

PVC Style Generation

Specially optimized for generating anime character images in PVC figure style

Multi-Trigger Word Support

Supports using trigger words like pvc, nendoroid, figma to generate different figure styles

High-Quality Output

Generated images are of masterpiece quality with optimal quality ratings

LoRA Fine-tuning

Utilizes Low-Rank Adaptation (LoRA) technology for efficient fine-tuning of the base model

Model Capabilities

Text-to-Image Generation

Anime Character Generation

PVC Figure Style Conversion

High-Quality Image Generation

Use Cases

Digital Art Creation

Anime Figure Concept Design

Design concept art for anime characters in PVC figure style

Generates high-quality figure-style images

Game Development

Game Character Figure Conversion

Convert game characters into figure style for merchandise design

Generates figure-style images suitable for commercial products

🚀 PVC Style LoRA for AnimagineXL 4.0 opt

This is a LoRA model for AnimagineXL 4.0 opt, which can generate PVC - style images. It has been tested with ComfyUI.

🚀 Quick Start

Model Gallery

✨ Features

Model description

This model has only been tested with ComfyUI.

Trigger words

Use pvc to trigger the image generation.
Use nendoroid to trigger the image generation.
Use figma to trigger the image generation.

📦 Installation

Download model

The weights for this model are available in Safetensors format. Download them in the Files & versions tab.

📚 Documentation

Dataset

The dataset consists of about 7000 PVC figure images.

Training

Training code

The training code can be found at https://github.com/p1atdev/vision - ft.

Training settings

`config.yml`

model:
  checkpoint_path: "./models/animagine-xl-4.0-opt.bnb_nf4.safetensors"
  pretrained_model_name_or_path: cagliostrolab/animagine-xl-4.0

  dtype: bfloat16

  denoiser:
    attn_implementation: "flash_attention_2"

peft:
  config:
    type: lora
    rank: 2
    alpha: 1.0
    dropout: 0.0

    dtype: bfloat16

  include_keys:
    - "attn1"
    - "attn2"
    - ".ff."
  exclude_keys: ["text_encoder", "vae"]

dataset:
  folder: "data/pvc"
  num_repeats: 4
  batch_size: 4

  bucket_base_size: 1024
  step: 128
  min_size: 384
  do_upscale: true

  caption_processors:
    - type: shuffle
      split_separator: ","

optimizer:
  name: "schedulefree.RAdamScheduleFree"
  args:
    lr: 0.005

scheduler:
  # name: "torch.optim.lr_scheduler.ConstantLR"
  # args: {}

tracker:
  project_name: "sdxl-pvc-1"
  loggers:
    - wandb

saving:
  strategy:
    per_epochs: 0.25
    per_steps: null
    save_last: true

  callbacks:
    - type: "safentensors"
      name: "sdxl-pvc"
      save_dir: "./output/sdxl-pvc-2"

preview:
  strategy:
    per_epochs: 1
    per_steps: 100

  callbacks:
    - type: "local"
      save_dir: "./output/sdxl-lora/preview"

  data:
    path: "./projects/pvc-sdxl/preview.yml"

seed: 42
num_train_epochs: 10

trainer:
  # debug_mode: "1step"

  gradient_checkpointing: true

  torch_compile: true
  torch_compile_args:
    backend: eager
    mode: default
    fullgraph: false

  fp32_matmul_precision: "high"
  allow_tf32: true

`preview.yml`

- prompt: |-
    pvc style, 1girl, aqua eyes, baseball cap, blonde hair, closed mouth, earrings, 
    green background, hat, hoop earrings, jewelry, looking at viewer, 
    shirt, short hair, simple background, solo, upper body, yellow shirt, 
    masterpiece, high score, great score, absurdres
  negative_prompt: |-
    lowres, bad anatomy, bad hands, text, error, missing finger, extra digits, 
    fewer digits, cropped, worst quality, low quality, low score, bad score, 
    average score, signature, watermark, username, blurry
  height: 1024
  width: 1024
  cfg_scale: 5.0
  num_steps: 20

  seed: 0

The training was stopped in the 9th epoch because the room was too hot due to this training.

Device and Hours

The model was trained on an RTX 4070 Ti Super (VRAM 16GB) for 27 hours.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご