- Download the zip file containing the custom nodes.
- Extract the files to the following directory:
..\ComfyUI\custom_nodes
- Restart ComfyUI to load the new nodes.
A complete workflow demonstrating the usage of this extension is available here:
ComfyUI-In-Context-LoRA Workflow
To make the workflow functional, ensure the following files are downloaded and placed in the appropriate directories:
-
flux1-fill-dev.safetensors
Download here
Place in:..\ComfyUI\ComfyUI\models\unet
-
ViT-L-14-TEXT-detail-improved-hiT-GmP-HF.safetensors
Download here
Place in:..\ComfyUI\ComfyUI\models\clip
-
t5xxl_fp8_e4m3fn_scaled.safetensors
Download here
Place in:..\ComfyUI\ComfyUI\models\clip
-
ae.sft
Download here
Place in:..\ComfyUI\models\vae
-
flux1-redux-dev.safetensors
Download here
Place in:..\ComfyUI\models\style_models
-
sigclip_vision_patch14_384.safetensors
Download here
Place in:..\ComfyUI\models\clip_vision
- 2024-12-14: Adjust x_diff calculation and adjust fit image logic.
- 2024-12-13: Fix Incorrect Padding
- 2024-12-12(2): Fix center point calculation when close to edge.
- 2024-12-12: Reconstruct the node with new caculation.
- 2024-12-11: Avoid too large buffer cause incorrect context area
- 2024-12-10(3): Avoid padding when image have width or height to extend the context area
- 2024-12-05: Fix incorrect cropping issue when mask is not fit in target ratio
- 2024-12-01: Adjust contours calculation, fix multiple contour only produce single contour issue
- 2024-11-30: Add AutoPatch node, it is able to automatically select patch mode and patch type
- 2024-11-29: Recontruct the node and seperate from old node, new nodes: CreateContextWindow, ConcatContextWindow
- 2024-11-22: Update Two Images input and related masks input
Simple Try On Lora: https://civitai.com/models/950111/flux-simple-try-on-in-context-lora
- Twitter: @Lrzjason
- Email: [email protected]
- QQ Group: 866612947
- Civitai: xiaozhijason