Instructions to use navervision/CompoDiff-Aesthetic with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use navervision/CompoDiff-Aesthetic with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("navervision/CompoDiff-Aesthetic", device_map="auto") - Notebooks
- Google Colab
- Kaggle
| { | |
| "architectures": [ | |
| "CompoDiffModel" | |
| ], | |
| "embed_dim": 768, | |
| "model_depth": 12, | |
| "model_dim": 64, | |
| "model_heads": 16, | |
| "model_type": "CompoDiff", | |
| "timesteps": 1000, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.25.1" | |
| } | |