---
license: mit
library_name: diffusers
pipeline_tag: text-to-image
tags:
- text-to-image
- diffusers
- clip
- sprited
datasets:
- sprited/dancing-chibi-figures
---

# dancing-chibi-figures-t2i-64

**Why this model exists:** it is the finished version of the model the
[Colab tutorial](https://github.com/sprited-ai/dancing-chibi-figures/blob/main/notebooks/dancing_chibi_figures_colab.ipynb)
has you build — a **miniature Stable Diffusion**: the same standard parts (`UNet2DConditionModel` + a frozen CLIP
ViT-B/32 text encoder + classifier-free guidance), minus the VAE, diffusing 64px RGBA pixels directly. It is the first
model in the [dancing-chibi-figures](https://huggingface.co/datasets/sprited/dancing-chibi-figures) family that reads
the actual sentence (via cross-attention) rather than a class label. 90M UNet params, 30k steps, EMA.

<p align="center"><img src="sample_030000.png" width="512"></p>

*One column per prompt: jump / dance / cartwheel / kneel / walk / raise hands / kick / sit.*

```python
from diffusers import DiffusionPipeline
pipe = DiffusionPipeline.from_pretrained("sprited/dancing-chibi-figures-t2i-64",
                                         custom_pipeline="sprited/dancing-chibi-figures-t2i-64",
                                         trust_remote_code=True).to("cuda")
imgs = pipe(["a person jumps in place"] * 4, num_inference_steps=50, guidance_scale=3.0).images  # RGBA PIL
```

`guidance_scale=0` ignores the text entirely; ~3 is a good default.

## Honest limitations

- Trained on **143 unique motion prompts**. A novel sentence is pulled toward the most similar trained one — this is
  closer to "knows its 143 moves by name" than to open-vocabulary understanding. (More prompt diversity is the v0.2 plan.)
- Prompts describe a whole **motion**, not one pose: "does a cartwheel" legitimately samples any split-second of a
  cartwheel, so the same prompt yields different poses.
- Structural check (silhouette oracle v0, n=256 vs 512 real): one_piece 100% / stray 0.0% / fg 0.116 — identical to the real floor (100% / 0.0% / 0.119). Prompt adherence is not yet scored
  systematically.

Training script: [`train_diffusers/train_t2i.py`](https://github.com/sprited-ai/dancing-chibi-figures/blob/main/train_diffusers/train_t2i.py).
Siblings: [ddpm-64](https://huggingface.co/sprited/dancing-chibi-figures-ddpm-64) (unconditional diffusers) ·
[baselines](https://huggingface.co/sprited/dancing-chibi-figures-baselines) (pure-PyTorch, incl. video).
Made by [Sprited](https://sprited.ai).