ModelMergeCosmos7B - ComfyUI Built-in Node Documentation - ComfyUI

The ModelMergeCosmos7B node merges two AI models together using weighted blending of specific components. It allows fine-grained control over how different parts of the models are combined by adjusting individual weights for position embeddings, transformer blocks, and final layers.

Inputs

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

All weight parameters accept values from 0.0 to 1.0 in steps of 0.01 and default to 1.0.

Outputs

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

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

Source fingerprint (SHA-256): 2cc4dcaa3576c5383c630e233cef55dedc8d742c20197cc83f5832dc9e887dac