stable-diffusion-1-5-neco-arc-zxc

abbyrana/stable-diffusion-1-5-neco-arc-zxc

A fine-tuned Stable Diffusion v1.5 for generating images of Neco Arc using the trigger token zxc.

---

Generated Images

<div style="display:flex; flex-wrap:wrap; gap:10px;"> <div style="flex:1 1 200px"> <img src="results/zxc as a solider.jpeg" alt="zxc as a solider" width="100%"> <pre><code>zxc as a solider</code></pre> </div> <div style="flex:1 1 200px"> <img src="results/zxc in a bucket.jpeg" alt="zxc in a bucket" width="100%"> <pre><code>zxc in a bucket</code></pre> </div> <div style="flex:1 1 200px"> <img src="results/zxc wearing a crown.jpeg" alt="zxc wearing a crown" width="100%"> <pre><code>zxc wearing a crown</code></pre> </div> <div style="flex:1 1 200px"> <img src="results/zxc wearing a tuxedo.jpeg" alt="zxc wearing a tuxedo" width="100%"> <pre><code>zxc wearing a tuxedo</code></pre> </div> <div style="flex:1 1 200px"> <img src="results/zxc wearing spiderman suit.jpeg" alt="zxc wearing spiderman suit" width="100%"> <pre><code>zxc wearing spiderman suit</code></pre> </div> </div>

---

Usage

from diffusers import StableDiffusionPipeline
import torch
from IPython.display import display

pipe = StableDiffusionPipeline.from_pretrained(
    "abbyrana/stable-diffusion-1-5-neco-arc-zxc",
    torch_dtype=torch.float16
).to("cuda")

# Recommended: 512x512 resolution, 100 steps
image = pipe(
    "zxc in a bucket", # <- Your prompt
    height=512,
    width=512,
    num_inference_steps=100
).images[0]

# Display inline
display(image)

# Save locally
image.save("neco_arc.png")

Prompt examples:

zxc inside a jar, anime style

zxc wearing a wizard hat, highly detailed

---

Limitations & Safety

• Optimized for anime / 3D render / plush-style outputs
• Not suitable for realistic humans
• Do not use for NSFW, illegal, or harmful content

---

Training

• Method: Fast DreamBooth
• Dataset: Neco Arc images with English captions
• Time: ~2 hours
• Hardware: Google Colab (T4 GPU)

---

Environmental Impact

Emissions can be estimated using the ML CO₂ Impact Calculator.

---

Citation

If you have found this work valuable, please consider citing it:

@misc{abbyrana2026necoarc,
  title = {abbyrana/stable-diffusion-1-5-neco-arc-zxc},
  author = {Abhijit Rana},
  year = {2026},
  howpublished = {\url{https://huggingface.co/abbyrana/stable-diffusion-1-5-neco-arc-zxc}},
  note = {If you find this work valuable, your citation and support are appreciated}
}