Nodes/ComfyUI-TinyBreaker/💪TB | Tiny Dual Sampler
ComfyUI Node

💪TB | Tiny Dual Sampler

Denoises the latent image in two stages: first with a base model, then with a refiner model to add details and enhance image quality.

By martin-rizzo·Created 2 years ago·Updated 12 months ago· 42
💪TB | Tiny Dual Sampler
  • latent_input
  • genparams
  • model
  • clip
  • transcoder
  • refiner_model
  • refiner_clip
  • latent_output
enable_cachetrue
Category💪TinyBreaker

Inputs (8)

NameTypeDefaultDescription
latent_inputLATENTThe latent image to denoise.
genparamsGENPARAMSThe generation parameters containing the sampler configuration.
modelMODELThe model used for denoising the latent images.
clipCLIPThe T5 encoder used for embedding the prompts.
transcoderTRANSCODERThe transcoder model used for converting latent images from base to refiner.
refiner_modelMODELThe model used for refining latent images.
refiner_clipCLIPThe CLIP model used for embedding text prompts during refining.
enable_cacheBOOLEANtrueEnable internal cache to avoid redundant computations. It might be necessary to disable it if errors occur or the image is not updating correctly.

Outputs (1)

NameTypeDescription
latent_outputLATENTLatent image after denoising.