LoadTrainingDataset - ComfyUI Built-in Node Documentation

Complete documentation for the LoadTrainingDataset node in ComfyUI. Learn its inputs, outputs, parameters and usage.

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

This node loads an encoded training dataset that has been previously saved to disk. It searches for and reads all data shard files from a specified folder within the ComfyUI output directory, then returns the combined latent vectors and conditioning data for use in training workflows.

Inputs

Parameter	Data Type	Required	Range	Description
folder_name	STRING	No	N/A	Name of the folder containing the saved dataset, located inside the ComfyUI output directory (default: “training_dataset”).

Outputs

Output Name	Data Type	Description
latents	LATENT	A list of latent dictionaries, where each dictionary contains a "samples" key with a tensor.
conditioning	CONDITIONING	A list of conditioning lists, where each inner list contains conditioning data for a corresponding sample.