rllm.llm.FeatLLM¶
- class rllm.llm.FeatLLM(file_path: str, metadata_path: str, task_info_path: str, llm: BaseChatModel | BaseLLM | None = None, **kwargs)[source]¶
Bases:
objectA class that integrates feature engineering and LLM-based tasks.
- feat_engineer¶
An instance of the feature engineering class.
- Type:
- kwargs¶
Additional arguments for customization.
- Type:
dict
- invoke()[source]¶
Executes the feature engineering and model training pipeline.
This method performs the following steps: 1. Generates features using the feature engineer. 2. Trains a model for each executable feature set. 3. Evaluates the model and computes AUC scores. 4. Ensembles the results and computes the final AUC score.