Skip to main content
This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub
The LotusConditioning node provides pre-computed conditioning embeddings for the Lotus model. It uses a frozen encoder with null conditioning and returns hardcoded prompt embeddings to achieve parity with the reference implementation without requiring inference or loading large tensor files. This node outputs a fixed conditioning tensor that can be used directly in the generation pipeline.

Inputs

ParameterData TypeRequiredRangeDescription
No inputs---This node does not accept any input parameters.

Outputs

Output NameData TypeDescription
conditioningCONDITIONINGThe pre-computed conditioning embeddings for the Lotus model, containing fixed prompt embeddings and an empty dictionary.