File size: 1,748 Bytes
4241cb3
 
 
22426be
 
 
 
 
 
 
 
 
752a0b4
22426be
 
 
 
9567af1
 
 
 
 
 
 
 
 
752a0b4
9567af1
22426be
 
 
 
 
 
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
---
license: creativeml-openrail-m
---

## Usage

```python
from diffusers import StableDiffusionPipeline

pipe = StableDiffusionPipeline.from_pretrained("lambdalabs/miniSD-diffusers")
pipe = pipe.to("cuda")

prompt = "a photograph of an wrinkly old man laughing"
image = pipe(prompt, width=256, height=256).images[0]  
image.save('test.jpg')
```

## Training details

Fine tuned from the stable-diffusion 1.4 checkpoint as follows:

- 22,000 steps fine-tuning only the attention layers of the unet, learn rate=1e-5, batch size=256
- 66,000 steps training the full unet, learn rate=5e-5, batch size=552

- GPUs provided by [Lambda GPU Cloud](https://lambdalabs.com/service/gpu-cloud)
- Trained on [LAION Improved Aesthetics 6plus](https://huggingface.co/datasets/ChristophSchuhmann/improved_aesthetics_6plus).
- Trained using https://github.com/justinpinkney/stable-diffusion, original [checkpoint available here](https://huggingface.co/justinpinkney/miniSD)

## License
This model is open access and available to all, with a CreativeML OpenRAIL-M license further specifying rights and usage. The CreativeML OpenRAIL License specifies:

- You can't use the model to deliberately produce nor share illegal or harmful outputs or content
- The authors claims no rights on the outputs you generate, you are free to use them and are accountable for their use which must not go against the provisions set in the license
- You may re-distribute the weights and use the model commercially and/or as a service. If you do, please be aware you have to include the same use restrictions as the ones in the license and share a copy of the CreativeML OpenRAIL-M to all your users (please read the license entirely and carefully) Please read the full license here