Skip to main content
This node merges two AI models using a block-based approach with fine-grained control over different model components. It allows you to blend models by adjusting interpolation weights for specific sections including positional frequencies, embedding layers, and individual transformer blocks. The merging process combines the architectures and parameters from both input models according to the specified weight values.

Inputs

Outputs

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): aebf536f3f89ca8c81141ac871b1b612082c3bd38a29984168b05eccf0cb57e3