|
--- |
|
license: mit |
|
metrics: |
|
- mse |
|
library_name: diffusers |
|
tags: |
|
- diffusion |
|
pipeline_tag: unconditional-image-generation |
|
--- |
|
|
|
## Obama Model Card |
|
|
|
DDPMObama is a latent noise-to-image diffusion model capable of generating images of obama. For more information about how Stable Diffusion functions, please have a look at 🤗's [Stable Diffusion blog](https://huggingface.co/blog/stable_diffusion). |
|
|
|
You can use this with the 🧨Diffusers library from [Hugging Face](https://huggingface.co). |
|
|
|
 |
|
|
|
### Diffusers |
|
```py |
|
from diffusers import DiffusionPipeline |
|
|
|
pipeline = DiffusionPipeline.from_pretrained("nroggendorff/obama") |
|
pipe = pipeline.to("cuda") |
|
|
|
image = pipe().images[0] |
|
|
|
image.save("obama.png") |
|
``` |
|
|
|
|
|
### Model Details |
|
- `train_batch_size`: 16 |
|
- `eval_batch_size`: 16 |
|
- `num_epochs`: 50 |
|
- `gradient_accumulation_steps`: 1 |
|
- `learning_rate`: 1e-4 |
|
- `lr_warmup_steps`: 500 |
|
- `mixed_precision`: "fp16" |
|
- `eval_metric`: "mean_squared_error" |
|
|
|
### Limitations |
|
|
|
- The model does not achieve perfect photorealism |
|
- The model cannot render legible text |
|
- The model was trained on a medium-to-large-scale dataset: [few-shot-obama](https://huggingface.co/datasets/huggan/few-shot-obama) |
|
|
|
### Developed by |
|
- Noa Linden Roggendorff |
|
|
|
*This model card was written by Noa Roggendorff and is based on the [Stable Diffusion v1-5 Model Card](https://huggingface.co/runwayml/stable-diffusion-v1-5).* |
