As the whole document is just too huge to suit in to the context window of the LLM, you will need to partition it into scaled-down text chunks, which can be termed Nodes in LlamaIndex. you could parse the loaded files https://lucyfdbo300043.wikicorrespondence.com/user