Split the text into smaller pieces that can be read by a large language model, since those models have limits to how much they can read at once. The 300 pages of text will exceed OpenAI’s limits.

Phi-2 is a generative AI model with 2.7 billion-parameters used for research and development of language models.