camenduru/dynami-crafter 🔢📝🖼️ → 🖼️

▶️ 6.4K runs 📅 Dec 2023 ⚙️ Cog 0.9.0-beta11 🔗 GitHub 📄 Paper ⚖️ License

animation image-to-video motion-control video-diffusion

About

DynamiCrafter: Animating Open-domain Images with Video Diffusion Priors

Example Output

Output

Performance Metrics

34.17s Prediction Time

185.39s Total Time

All Input Parameters

{
  "i2v_eta": 1,
  "i2v_seed": 123,
  "i2v_steps": 50,
  "i2v_motion": 3,
  "i2v_cfg_scale": 7.5,
  "i2v_input_text": "boy walking on the street",
  "i2v_input_image": "https://replicate.delivery/pbxt/K1vHFesvAE0TPLLLY5J5KFFWBVRSh5Akzy6E5oPHvOqpDDGF/boy.webp"
}

Input Parameters

i2v_eta Type: numberDefault: 1Range: 0 - 1: ETA
i2v_seed Type: integerDefault: 123Range: 0 - 10000: Random Seed
i2v_steps Type: integerDefault: 50Range: 1 - 60: Sampling steps
i2v_motion Type: integerDefault: 3Range: 1 - 4: Motion magnitude
i2v_cfg_scale Type: numberDefault: 7.5Range: 1 - 15: CFG Scale
i2v_input_text Type: stringDefault: man fishing in a boat at sunset
i2v_input_image (required) Type: string: Input image

Output Schema

Output

Type: string • Format: uri

Example Execution Logs

Seed set to 123
start: boy walking on the street 2023-12-11 12:18:41
/usr/local/lib/python3.10/site-packages/torchvision/transforms/functional.py:1603: UserWarning: The default value of the antialias parameter of all the resizing transforms (Resize(), RandomResizedCrop(), etc.) will change from None to True in v0.17, in order to be consistent across the PIL and Tensor backends. To suppress this warning, directly pass antialias=True (recommended, future default), antialias=None (current default, which means False for Tensors and True for PIL), or antialias=False (only works on Tensors - PIL will still use antialiasing). This also applies if you are using the inference transforms from the models weights: update the call to weights.transforms(antialias=True).
warnings.warn(
/usr/local/lib/python3.10/site-packages/torch/nn/modules/conv.py:459: UserWarning: Applied workaround for CuDNN issue, install nvrtc.so (Triggered internally at ../aten/src/ATen/native/cudnn/Conv_v8.cpp:80.)
return F.conv2d(input, weight, bias, self.stride,

Version Details

Version ID: 17daec0f8d078e021ba8997118643d0a455d8385f92c44d262e5ce8c259d0575
Version Created: December 11, 2023

Run on Replicate →