File size: 2,634 Bytes
1549a78
 
 
 
 
 
 
 
 
 
 
 
 
 
bd47fb0
1549a78
 
 
 
 
 
 
 
 
 
 
 
7446879
1549a78
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
---
license: mit
base_model:
- Shitao/OmniGen-v1
pipeline_tag: text-to-image
tags:
- image-to-image
---


This repo contains bitsandbytes 4bit-NF4 float16 model weights for [OmniGen-v1](https://huggingface.co/Shitao/OmniGen-v1). These are intended for Google Colab users or those with a GPU that does not support bfloat16. Other 4-bit seekers should prefer the [bf16-bnb-4bit](https://huggingface.co/gryan/OmniGen-v1-bnb-4bit) model as it produces higher quality images. For info about OmniGen see the [original model card](https://huggingface.co/Shitao/OmniGen-v1).


- 8-bit weights: [gryan/OmniGen-v1-bnb-8bit](https://huggingface.co/gryan/OmniGen-v1-bnb-8bit)
- 4-bit (bf16, nf4) weights: [gryan/OmniGen-v1-bnb-4bit](https://huggingface.co/gryan/OmniGen-v1-bnb-4bit)


## Usage
Set up your environment by following the original [Quick Start Guide](https://huggingface.co/Shitao/OmniGen-v1#5-quick-start) before getting started.

> [!IMPORTANT]
> NOTE: This feature is not officially supported yet. You'll need to install the repo from [this pull request](https://github.com/VectorSpaceLab/OmniGen/pull/151).

```python
from OmniGen import OmniGenPipeline, OmniGen

# pass the quantized model in the pipeline
model = OmniGen.from_pretrained('gryan/OmniGen-v1-fp16-bnb-4bit', dtype=torch.float16)
pipe = OmniGenPipeline.from_pretrained("Shitao/OmniGen-v1", model=model)

# proceed as normal!

## Text to Image
images = pipe(
    prompt="A curly-haired man in a red shirt is drinking tea.", 
    height=1024, 
    width=1024, 
    guidance_scale=2.5,
    seed=0,
)
images[0].save("example_t2i.png")  # save output PIL Image

## Multi-modal to Image
# In the prompt, we use the placeholder to represent the image. The image placeholder should be in the format of <img><|image_*|></img>
# You can add multiple images in the input_images. Please ensure that each image has its placeholder. For example, for the list input_images [img1_path, img2_path], the prompt needs to have two placeholders: <img><|image_1|></img>, <img><|image_2|></img>.
images = pipe(
    prompt="A man in a black shirt is reading a book. The man is the right man in <img><|image_1|></img>.",
    input_images=["./imgs/test_cases/two_man.jpg"],
    height=1024, 
    width=1024,
    guidance_scale=2.5, 
    img_guidance_scale=1.6,
    seed=0
)
images[0].save("example_ti2i.png")  # save output PIL image
```

## Image Samples
<img src="./assets/text_only_1111_fp16_4bit.png" alt="Text Only FP16 4bit">
<img src="./assets/single_img_1111_fp16_4bit.png" alt="Single Image FP16 4bit">
<img src="./assets/double_img_1111_fp16_4bit.png" alt="Double Image FP16 4bit">