A high-fidelity LoRA capturing the iconic, beloved look of the classic 1970s Soviet stop-motion puppet Cheburashka (Чебурашка). Trained on a clean, consistent dataset to recreate the authentic dark brown felt wool texture, large round ears, warm expressive button eyes, and vintage nostalgic aesthetic.### ⚙️ Recommended Settings &amp; Usage * Trigger Word: <code>cheburashka_classic</code> * Additional Style Boosters: <code>stop-motion puppet</code>, <code>felt wool texture</code>, <code>vintage 1970s aesthetic</code>, <code>round ears</code> * LoRA Weight (Standalone): <code>0.85 – 1.0</code> * LoRA Weight (Multi-LoRA stack with characters/vehicles): <code>0.55 – 0.65</code> (keep CLIP weight around <code>0.5 – 0.6</code> to avoid overbaking faces/clothing)cheburashka_classic, majestic portrait, vertical 2:3, head and shoulders, looking directly into camera with giant expressive glossy eyes, soft outdoor cinematic rim light, golden hour backlight creating warm halo around dense fur, rich textures, award-winning photography, 8kcheburashka_classic, vertical 2:3, full body, peeking curiously out of giant cracked cardboard delivery box overflowing with vibrant bright oranges, studio backdrop, dynamic lighting, high contrast, crisp sharp focuscheburashka_classic, vertical 2:3, full body, standing on hood of vintage weathered green soviet off-road truck, proud heroic pose, muddy country road, cinematic wide shot, detailed vintage atmosphere, 8kcheburashka_classic, vertical 2:3, full body, walking through dense colorful autumn park carrying tiny rustic lantern with warm glowing candle inside, mystical twilight atmosphere, falling orange leaves, cinematic depth of field

Cheburashka [MOSS / Krea 2\,Qwen 2.1]

cheburashkaV1Q.safetensors

v1.0 Qwen 2.1

v1.0 MOSS

v1.0 Krea 2


# 

## Model Introduction

This model is trained using the [Civision/model training](https://modelscope.ai/civision/modelTraining) environment and computing resources provided by ModelScope.

* Model Type: LoRA
* Base Model: [Qwen/Qwen-Image-2.1](https://modelscope.ai/models/Qwen/Qwen-Image-2.1)
* Training Code: [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio)
* Data amount：15
* Total training steps：2000
* License: Apache-2.0

## Inference Code

Install [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio):

```bash
pip install diffsynth
```

Inference:

```python
from diffsynth.pipelines.qwen_image_21 import QwenImage21Pipeline, ModelConfig
import torch

pipe = QwenImage21Pipeline.from_pretrained(
    torch_dtype=torch.bfloat16,
    device="cuda",
    model_configs=[
        ModelConfig(model_id="Qwen/Qwen-Image-2.1", origin_file_pattern="transformer/diffusion_pytorch_model*.safetensors"),
        ModelConfig(model_id="Qwen/Qwen-Image-2.1", origin_file_pattern="text_encoder/model*.safetensors"),
        ModelConfig(model_id="Qwen/Qwen-Image-2.1", origin_file_pattern="vae/diffusion_pytorch_model*.safetensors"),
    ],
    processor_config=ModelConfig(model_id="Qwen/Qwen-Image-2.1", origin_file_pattern="processor/"),
)
pipe.load_lora(pipe.dit, ModelConfig(model_id="bazya75/cheburashkaV1Q", origin_file_pattern="cheburashkaV1Q_c1-st2000.safetensors"))

prompt = "a cat"
image = pipe(prompt, seed=0, num_inference_steps=50)
image.save("image.png")
```


Files with SHA256: 813ce6aad9c4005cb8dc0c329d7c7fdd47c95a9c5c6530a5b53a51f4ee1bd6f5