unCLIPCheckpointLoader - ComfyUI Built-in Node Documentation

This node will detect models located in the ComfyUI/models/checkpoints folder, and it will also read models from additional paths configured in the extra_model_paths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. The unCLIPCheckpointLoader node is designed for loading checkpoints specifically tailored for unCLIP models. It facilitates the retrieval and initialization of models, CLIP vision modules, and VAEs from a specified checkpoint, streamlining the setup process for further operations or analyses.

Inputs

Field	Description	Comfy dtype
`ckpt_name`	Specifies the name of the checkpoint to be loaded, identifying and retrieving the correct checkpoint file from a predefined directory, determining the initialization of models and configurations.	`COMBO[STRING]`

Outputs

Field	Description	Comfy dtype	Python dtype
`model`	Represents the primary model loaded from the checkpoint.	`MODEL`	`torch.nn.Module`
`clip`	Represents the CLIP module loaded from the checkpoint, if available.	`CLIP`	`torch.nn.Module`
`vae`	Represents the VAE module loaded from the checkpoint, if available.	`VAE`	`torch.nn.Module`
`clip_vision`	Represents the CLIP vision module loaded from the checkpoint, if available.	`CLIP_VISION`	`torch.nn.Module`

This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

TripleCLIPLoader - ComfyUI Built-in Node Documentation

UNETLoader - ComfyUI Built-in Node Documentation

​Inputs

​Outputs

Inputs

Outputs