Instructions to use impactframes/Video-to-Video with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- OpenCLIP
How to use impactframes/Video-to-Video with OpenCLIP:
import open_clip model, preprocess_train, preprocess_val = open_clip.create_model_and_transforms('hf-hub:impactframes/Video-to-Video') tokenizer = open_clip.get_tokenizer('hf-hub:impactframes/Video-to-Video') - Notebooks
- Google Colab
- Kaggle
| { | |
| "framework": "pytorch", | |
| "task": "video-to-video", | |
| "model": { | |
| "type": "video-to-video-model", | |
| "model_args": { | |
| "ckpt_clip": "open_clip_pytorch_model.bin", | |
| "ckpt_unet": "non_ema_0035000.pth", | |
| "ckpt_autoencoder": "v2-1_512-ema-pruned.ckpt", | |
| "seed": 666, | |
| "solver_mode": "fast" | |
| }, | |
| "model_cfg": { | |
| "batch_size": 1, | |
| "target_fps": 8, | |
| "max_frames": 32, | |
| "latent_hei": 90, | |
| "latent_wid": 160 | |
| } | |
| }, | |
| "pipeline": { | |
| "type": "video-to-video-pipeline" | |
| } | |
| } | |