search

energy-based-model / Compositional-Visual-Generation-with-Composable-Diffusion-Models-PyTorch

[ECCV 2022] Compositional Generation using Diffusion Models
https://energy-based-model.github.io/Compositional-Visual-Generation-with-Composable-Diffusion-Models/
Other
456 stars 41 forks source link

minor fix in inference command #10

Closed tejas-gokhale closed 2 years ago

tejas-gokhale commented 2 years ago

https://github.com/energy-based-model/Compositional-Visual-Generation-with-Composable-Diffusion-Models-PyTorch#python

The prompt should be put in quotes. python scripts/image_sample_compose_glide.py --prompt "a camel | a forest"

Also the commands for GLIDE and Stable-Diffusion should be swapped.

Thanks! Great work!

nanlliu commented 2 years ago

thanks for pointing that out! just fixed.

tejas-gokhale commented 2 years ago

another point:

in https://github.com/energy-based-model/Compositional-Visual-Generation-with-Composable-Diffusion-Models-PyTorch/blob/d047d034da5ab42e54137f2ce2c05fd78e2b220f/scripts/image_sample_compose_stable_diffusion.py#L14

add default variable types (Otherwise it throws errors)

parser.add_argument("--steps", default=50, type=int)
parser.add_argument("--scale", default=10, type=float)

You can close the issue after updating it! Thanks!

nanlliu commented 2 years ago

thanks! fixed.