Update all files for BitDance-14B-16x-diffusers
Browse files
README.md
ADDED
|
@@ -0,0 +1,75 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
library_name: diffusers
|
| 4 |
+
pipeline_tag: text-to-image
|
| 5 |
+
base_model: shallowdream204/BitDance-14B-16x
|
| 6 |
+
language:
|
| 7 |
+
- en
|
| 8 |
+
tags:
|
| 9 |
+
- bitdance
|
| 10 |
+
- text-to-image
|
| 11 |
+
- custom-pipeline
|
| 12 |
+
- diffusers
|
| 13 |
+
- qwen
|
| 14 |
+
---
|
| 15 |
+
|
| 16 |
+
# BitDance-14B-16x (Diffusers)
|
| 17 |
+
|
| 18 |
+
Diffusers-converted checkpoint for BitDance-14B-16x with bundled custom pipeline code (`bitdance_diffusers`) so it can be loaded directly with `DiffusionPipeline`.
|
| 19 |
+
|
| 20 |
+
## Quickstart (native diffusers)
|
| 21 |
+
|
| 22 |
+
```python
|
| 23 |
+
import torch
|
| 24 |
+
from diffusers import DiffusionPipeline
|
| 25 |
+
|
| 26 |
+
repo_id = "BiliSakura/BitDance-14B-16x-diffusers"
|
| 27 |
+
|
| 28 |
+
pipe = DiffusionPipeline.from_pretrained(
|
| 29 |
+
repo_id,
|
| 30 |
+
trust_remote_code=True,
|
| 31 |
+
torch_dtype=torch.bfloat16,
|
| 32 |
+
).to("cuda")
|
| 33 |
+
|
| 34 |
+
result = pipe(
|
| 35 |
+
prompt="A cinematic landscape photo of snowy mountains at sunrise.",
|
| 36 |
+
height=1024,
|
| 37 |
+
width=1024,
|
| 38 |
+
num_inference_steps=50,
|
| 39 |
+
guidance_scale=7.5,
|
| 40 |
+
)
|
| 41 |
+
result.images[0].save("bitdance_14b_16x.png")
|
| 42 |
+
```
|
| 43 |
+
|
| 44 |
+
## Model Metadata
|
| 45 |
+
|
| 46 |
+
- Pipeline class: `BitDanceDiffusionPipeline`
|
| 47 |
+
- Diffusers version in config: `0.36.0`
|
| 48 |
+
- Parallel prediction factor: `16`
|
| 49 |
+
- Text stack: `Qwen3ForCausalLM` + `Qwen2TokenizerFast`
|
| 50 |
+
- Supported resolutions include `1024x1024`, `1280x768`, `768x1280`, `2048x512`, and more (see `model_index.json`)
|
| 51 |
+
|
| 52 |
+
## Citation
|
| 53 |
+
|
| 54 |
+
If you use this model, please cite BitDance and Diffusers:
|
| 55 |
+
|
| 56 |
+
```bibtex
|
| 57 |
+
@article{ai2026bitdance,
|
| 58 |
+
title = {BitDance: Scaling Autoregressive Generative Models with Binary Tokens},
|
| 59 |
+
author = {Ai, Yuang and Han, Jiaming and Zhuang, Shaobin and Hu, Xuefeng and Yang, Ziyan and Yang, Zhenheng and Huang, Huaibo and Yue, Xiangyu and Chen, Hao},
|
| 60 |
+
journal = {arXiv preprint arXiv:2602.14041},
|
| 61 |
+
year = {2026}
|
| 62 |
+
}
|
| 63 |
+
|
| 64 |
+
@inproceedings{von-platen-etal-2022-diffusers,
|
| 65 |
+
title = {Diffusers: State-of-the-art diffusion models},
|
| 66 |
+
author = {Patrick von Platen and Suraj Patil and Anton Lozhkov and Damar Jablonski and Hernan Bischof and Thomas Wolf},
|
| 67 |
+
booktitle = {GitHub repository},
|
| 68 |
+
year = {2022},
|
| 69 |
+
url = {https://github.com/huggingface/diffusers}
|
| 70 |
+
}
|
| 71 |
+
```
|
| 72 |
+
|
| 73 |
+
## License
|
| 74 |
+
|
| 75 |
+
This repository is distributed under the Apache-2.0 license, consistent with the upstream BitDance release.
|