ComfyUI Extension: ComfyUI LLM Promp
Run ComfyUI workflows without the setup
No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.
NODES: Load llm, Generate Text with LLM, Inference Qwen2VL, Inference Qwen2
Looking for a different extension?
Custom Nodes (0)
README
ComfyUI LLM Promp
LLM Prompt
Qwen7B を用いた V2T, I2T を自動で行うカスタムノード
Add LLM Save Dir
mkdir /LLM/
And write it in extra_model_pathes.yaml
Add LLM Save Path in
folder_pathes.py
folder_names_and_paths["LLM"] = ([os.path.join(models_dir, "LLM")], supported_pt_extensions)
AlignPOSE_KEYPOINTToReference
This module is a custom ComfyUI node that uses pose_keypoints_2d data obtained from OpenPose (or DWPose) to align reference images to the posture of an input image, specifically by scaling and shifting based on the neck position.
The pose keypoint information of the reference image is used to align it to the input image.
-
Supports the Body_25 format (OpenPose keypoint format)
-
Scales the reference image based on the neck keypoint as the anchor.
Demo
Origin Reference Pose ↓↓
Input Image Pose (Target) ↓↓
Use These Two Pose Image and Key Point to Align.
Aligned Image Pose ↓↓
...
Run ComfyUI workflows without the setup
No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.