Musubi Cache Text Embeds (Wan)

A ComfyUI node in musubi-tuner/wan/preprocess with 9 inputs and 3 outputs.

By jaimitoes·Created about a year ago·Updated 12 months ago· 61

◄dataset_config_toml—►

◄t5_name▾►

◄fp8_t5false►

◄devicecuda►

◄batch_size_override0►

◄num_workers_override0►

◄skip_existingfalse►

◄keep_cachetrue►

Categorymusubi-tuner/wan/preprocess

Inputs (9)

Name	Type	Default	Description
dataset_config_toml	STRING		—
trigger_in	*		—
t5_name	COMBO		Text Encoder (T5). Select the T5 model that your base model uses. This component is the "translator" that converts your prompts (text) into a language the image model understands.
fp8_t5	BOOLEAN	false	Enable to use 8-bit precision (FP8) for the T5 text encoder. Saves a lot of memory, but can be experimental and not always stable, and not all GPUs support it.
device	COMBO	cuda	Choose the hardware device to use for processing. `cuda` uses your NVIDIA GPU (much faster), while `cpu` uses your computer's main processor (much slower).
batch_size_override	INT	0	Override the default batch size used for processing data. `0` means use the default value from the dataset configuration. A higher number processes more items at once, but uses more memory.
num_workers_override	INT	0	Override the default number of workers for loading data. `0` means use the default value from the dataset configuration. A higher number can speed up data loading, but consumes more CPU and RAM.
skip_existing	BOOLEAN	false	If enabled, the node will skip processing files that already have existing cached versions, saving time and computation.
keep_cache	BOOLEAN	true	If enabled, the processed data will be kept in the cache directory even after this operation completes, so it can be reused later without reprocessing. If disabled, cached files might be removed.