RAGSynth

We will open-source the dataset and scripts after the review is finished.

1. wash_data_scripts: This section illustrates some of the data cleaning processes we employed during data collection.
2. src: This section differentiates between datasets, highlighting which datasets are used for evaluation and which serve as corpora.
3. knowledge_enhancement: This section demonstrates an implementation of RAGSynth.
   1. assemble: The process of synthesizing data using the content in the components.
   2. components: A specific implementation of RAGSynth in the form of a pipeline, where each step produces an output that serves as the input for the next step. This includes the synthesis of both single-hop and multi-hop data.
   3. chunk_by_files: The logic for chunking documents.
   4. rag：training scripts and logs；