| license: cc-by-nc-4.0 | |
| language: | |
| - en | |
| base_model: black-forest-labs/FLUX.1-Kontext-dev | |
| tags: | |
| - image-editing | |
| - image-retouching | |
| - lora | |
| - eyecontrol | |
| # EyeControl | |
| Checkpoints for **Dotting the Eye: An Intent-Driven Image Retouching | |
| Agent for Visual Focus Enhancement**. | |
| EyeControl supports instruction-based image retouching with global | |
| color and tone adjustments, local instructions, and spatial intent | |
| for visual focus enhancement. | |
| ## Files | |
| - `control.safetensors`: EyeControl control weights. | |
| - `adapter_model.safetensors`: trained LoRA adapter weights. | |
| - `adapter_config.json`: LoRA adapter configuration. | |
| These files require the FLUX.1-Kontext-dev base model and the | |
| EyeControl inference code. | |
| ## Download | |
| Run from the EyeControl code repository: | |
| ```bash | |
| hf download Dragoniss/EyeControl --local-dir pretrained/eyecontrol | |
| ``` | |
| ## Usage | |
| 1. Prepare FLUX.1-Kontext-dev at `pretrained/FLUX.1-Kontext-dev`. | |
| 2. Download the EyeControl checkpoint using the command above. | |
| 3. Set `CHECKPOINT_DIRS` in `infer.sh` to `("pretrained/eyecontrol")`. | |
| 4. Configure the input images, masks, and prompts, then run: | |
| ```bash | |
| bash infer.sh | |
| ``` |