ModelMergeCosmos14B - ComfyUI Built-in Node Documentation

Complete documentation for the ModelMergeCosmos14B 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

The ModelMergeCosmos14B node merges two AI models using a block-based approach specifically designed for the Cosmos 14B model architecture. It allows you to blend different components of the models by adjusting weight values between 0.0 and 1.0 for each model block and embedding layer.

Inputs

Parameter	Data Type	Required	Range	Description
model1	MODEL	Yes	-	First model to merge
model2	MODEL	Yes	-	Second model to merge
pos_embedder.	FLOAT	Yes	0.0 - 1.0	Weight for the position embedder component (default: 1.0)
extra_pos_embedder.	FLOAT	Yes	0.0 - 1.0	Weight for the extra position embedder component (default: 1.0)
x_embedder.	FLOAT	Yes	0.0 - 1.0	Weight for the x embedder component (default: 1.0)
t_embedder.	FLOAT	Yes	0.0 - 1.0	Weight for the t embedder component (default: 1.0)
affline_norm.	FLOAT	Yes	0.0 - 1.0	Weight for the affine normalization component (default: 1.0)
blocks.block0.	FLOAT	Yes	0.0 - 1.0	Weight for block 0 (default: 1.0)
blocks.block1.	FLOAT	Yes	0.0 - 1.0	Weight for block 1 (default: 1.0)
blocks.block2.	FLOAT	Yes	0.0 - 1.0	Weight for block 2 (default: 1.0)
blocks.block3.	FLOAT	Yes	0.0 - 1.0	Weight for block 3 (default: 1.0)
blocks.block4.	FLOAT	Yes	0.0 - 1.0	Weight for block 4 (default: 1.0)
blocks.block5.	FLOAT	Yes	0.0 - 1.0	Weight for block 5 (default: 1.0)
blocks.block6.	FLOAT	Yes	0.0 - 1.0	Weight for block 6 (default: 1.0)
blocks.block7.	FLOAT	Yes	0.0 - 1.0	Weight for block 7 (default: 1.0)
blocks.block8.	FLOAT	Yes	0.0 - 1.0	Weight for block 8 (default: 1.0)
blocks.block9.	FLOAT	Yes	0.0 - 1.0	Weight for block 9 (default: 1.0)
blocks.block10.	FLOAT	Yes	0.0 - 1.0	Weight for block 10 (default: 1.0)
blocks.block11.	FLOAT	Yes	0.0 - 1.0	Weight for block 11 (default: 1.0)
blocks.block12.	FLOAT	Yes	0.0 - 1.0	Weight for block 12 (default: 1.0)
blocks.block13.	FLOAT	Yes	0.0 - 1.0	Weight for block 13 (default: 1.0)
blocks.block14.	FLOAT	Yes	0.0 - 1.0	Weight for block 14 (default: 1.0)
blocks.block15.	FLOAT	Yes	0.0 - 1.0	Weight for block 15 (default: 1.0)
blocks.block16.	FLOAT	Yes	0.0 - 1.0	Weight for block 16 (default: 1.0)
blocks.block17.	FLOAT	Yes	0.0 - 1.0	Weight for block 17 (default: 1.0)
blocks.block18.	FLOAT	Yes	0.0 - 1.0	Weight for block 18 (default: 1.0)
blocks.block19.	FLOAT	Yes	0.0 - 1.0	Weight for block 19 (default: 1.0)
blocks.block20.	FLOAT	Yes	0.0 - 1.0	Weight for block 20 (default: 1.0)
blocks.block21.	FLOAT	Yes	0.0 - 1.0	Weight for block 21 (default: 1.0)
blocks.block22.	FLOAT	Yes	0.0 - 1.0	Weight for block 22 (default: 1.0)
blocks.block23.	FLOAT	Yes	0.0 - 1.0	Weight for block 23 (default: 1.0)
blocks.block24.	FLOAT	Yes	0.0 - 1.0	Weight for block 24 (default: 1.0)
blocks.block25.	FLOAT	Yes	0.0 - 1.0	Weight for block 25 (default: 1.0)
blocks.block26.	FLOAT	Yes	0.0 - 1.0	Weight for block 26 (default: 1.0)
blocks.block27.	FLOAT	Yes	0.0 - 1.0	Weight for block 27 (default: 1.0)
blocks.block28.	FLOAT	Yes	0.0 - 1.0	Weight for block 28 (default: 1.0)
blocks.block29.	FLOAT	Yes	0.0 - 1.0	Weight for block 29 (default: 1.0)
blocks.block30.	FLOAT	Yes	0.0 - 1.0	Weight for block 30 (default: 1.0)
blocks.block31.	FLOAT	Yes	0.0 - 1.0	Weight for block 31 (default: 1.0)
blocks.block32.	FLOAT	Yes	0.0 - 1.0	Weight for block 32 (default: 1.0)
blocks.block33.	FLOAT	Yes	0.0 - 1.0	Weight for block 33 (default: 1.0)
blocks.block34.	FLOAT	Yes	0.0 - 1.0	Weight for block 34 (default: 1.0)
blocks.block35.	FLOAT	Yes	0.0 - 1.0	Weight for block 35 (default: 1.0)
final_layer.	FLOAT	Yes	0.0 - 1.0	Weight for the final layer (default: 1.0)

Outputs

Output Name	Data Type	Description
model	MODEL	The merged model combining features from both input models

---

Source fingerprint (SHA-256): 6fcb4fefe7738d0addef49d386c0d3d22cda4c68f0e49ad003d1df595cf0e9d9