This changelog contains the notable updates to the RAG-Driven Generative AI repository.
🐬 Indicates new bonus notebooks to explore.
Implementing grok-beta to introduce RAG techniques:
Chapter01/RAG_Overview_Grok.ipynb
Chapter01/RAG_Overview.ipynb : removed import time because it is not necessary in this notebook.
nltk.download('punkt_tab') added to Chapter07/Wikipedia_API.ipynb
OpenAI Reasoning Models such as o1-preview can be used for RAG-Driven Generative AI in the ecosystem designed in Chapter 2.
🐬Chapter02/3_Augmented_Generation_o1_preview.ipynb
Fixed attribute error in the class RetrievalComponent: of the Modular RAG section in Chapter01/RAG_Overview.ipynb in Chapter 1.
self.documents is now initialized in the fit method to hold the records used for searching and enable the keyword_search function to access them without error.
def fit(self, records):
self.documents = records # Initialize self.documents here
if self.method == 'vector' or self.method == 'indexed':
self.tfidf_matrix = self.vectorizer.fit_transform(records)