Dumb shit to make life a lil more comfy
-
text_append_node.py Appends text to existing text, line by line.
-
vramdebugplus.py Monitors and manages VRAM usage.
- Optionally clears CUDA cache and performs garbage collection
- Can unload all models to free up VRAM
- Provides detailed or simple reports on VRAM usage
- Shows free VRAM before and after operations
- Offers insights into system RAM usage
- Useful for optimizing VRAM usage in complex ComfyUI workflows
-
tensordebugplus.py This script provides detailed information about tensors and neural network models.
- Analyzes the structure of tensors, including nested tensors and PyTorch modules
- Reports shape, data type, device, and memory usage of tensors
- Optionally includes gradient information and statistical measures (min, max, mean, std)
- Provides a summary of total parameters and memory usage
- Useful for debugging and optimizing machine learning models in ComfyUI
-
animation_schedule_output.py Designed to format text input for animation keyframes.
- Takes unformatted prompts and formats them with keyframe numbers
- Allows setting a custom keyframe interval
- Supports an optional offset for keyframe numbering
- Outputs formatted text suitable for animation schedules or keyframe-based systems
- Useful for preparing text inputs for animated or time-based ComfyUI workflows
These custom nodes extend the functionality of ComfyUI, providing tools for text manipulation, VRAM management, tensor analysis, and animation scheduling. They can be particularly useful for users working with complex workflows, large models, or animation projects in ComfyUI.
To use these tools, clone the repository to your local machine using:
cd Comfyui/custom_nodes
git clone https://github.com/KewkLW/ComfyUI-kewky_tools.git