adirik/stylemc 🖼️❓🔢📝 → 🖼️

▶️ 4.3K runs 📅 Mar 2024 ⚙️ Cog 0.9.4 🔗 GitHub 📄 Paper ⚖️ License

face-attribute-editing image-editing image-to-image

About

Text-guided image generation and editing

Example Output

Prompt:

"a photo of a happy face"

Output

Performance Metrics

50.61s Prediction Time

300.06s Total Time

All Input Parameters

{
  "image": "https://replicate.delivery/pbxt/KX8JZ1ZTBj6oSh78QKOExLXcDGWV4eB5Z3z9gnFCmMMJxynH/stylemc_sample.jpg",
  "prompt": "a photo of a happy face",
  "change_power": 8
}

Input Parameters

image (required) Type: string: Input face image.
prompt Default: a photo of a face with blonde hair: Choose a prompt with pre-calculated direction or specify the customized_prompt below
change_alpha Type: integerDefault: 10Range: -50 - 50: Strength coeffient to apply the style direction with.
custom_prompt Type: string: Specify your customized prompt. If set, the prompt above will be ignored. More descriptive texts produce better results such as using "a photo of a face with blonde hair" instead of "blonde hair".
id_loss_coeff Type: numberDefault: 2Range: 0 - 5: Identity loss coefficient, use higher values for better identity preservation.

Output Schema

Output

Type: string • Format: uri

Example Execution Logs

Setting up PyTorch plugin "bias_act_plugin"... Done.
Setting up PyTorch plugin "upfirdn2d_plugin"... Done.

Version Details

Version ID: 708f7bca82339e47999c6dcec9bae35e134997dc7cfa40108eea4e4f0defcfe7
Version Created: March 8, 2024

Run on Replicate →