

noobai-style,masterpiece,best quality,score_9,1girl,solo,white hair,blue eyes,hair ribbon,cute smile,looking at viewer,upper body,soft lighting,pastel background,detailed face

noobai-style,masterpiece,best quality,score_9,1girl,solo,pink hair,twintails,red eyes,sailor uniform,standing,full body,cherry blossoms,spring,outdoors,wind blowing hair,cute expression

noobai-style,masterpiece,best quality,score_9,1girl,solo,silver hair,cat ears,yellow eyes,sitting on bed,bedroom,warm lighting,holding stuffed animal,shy expression,detailed eyes

noobai-style,masterpiece,best quality,score_9,1girl,solo,blonde hair,side profile,looking back,white dress,flower field,sunset,golden hour,wind blowing,dreamy atmosphere

noobai-style,masterpiece,best quality,score_9,2girls,white hair girl,pink hair girl,school uniforms,holding hands,walking together,school corridor,sunlight through window,smiling

noobai-style,masterpiece,best quality,score_9,1girl,solo,close-up,face focus,green hair,big eyes,detailed eyes,soft lighting,blurred background,cute expression,looking at viewer
| Item | Value |
|---|---|
| Base Model | circlestone-labs/Anima-Base |
| LoRA Type | DiT LoRA (networks.lora_anima) |
| Network Dim | 16 |
| Network Alpha | 16 |
| Parameters | 1344 tensors |
| File Size | ~66 MB |
| Item | Value |
|---|---|
| Training Images | 123 |
| Epochs | 15 |
| Effective Steps | 465 |
| Batch Size | 1 |
| Gradient Accumulation | 4 |
| Learning Rate | 5e-5 |
| Optimizer | AdamW8bit (weight_decay=0.01) |
| LR Scheduler | Cosine |
| Warmup Steps | 100 |
| Mixed Precision | bf16 |
| Resolution | 1024×1024 (bucket-enabled) |
| Final Loss | ~0.108 |
noobai-style| Parameter | Value |
|---|---|
| Sampler | Euler a |
| Steps | 30 |
| CFG | 5.0 - 7.0 |
| Resolution | 1024×1024 |
noobai-style, 1girl, white hair, red eyes, standing in a flower field, wind blowing, masterpiece, best qualitynoobai-style, 1girl, cat ears, blue hair, sitting at a desk, reading a book, warm lighting, detailed background| File | Description |
|---|---|
noobai-style-anima-lora-final.safetensors | Final trained LoRA (recommended) |
noobai-style-anima-lora-000001.safetensors ~ 000014.safetensors | Per-epoch checkpoints |