Implementation

RAG from Scratch¶

Setup¶

In [ ]:

!pip install -q sentence-transformers
!pip install -q wikipedia-api
!pip install -q numpy
!pip install -q scipy

!pip install -q sentence-transformers !pip install -q wikipedia-api !pip install -q numpy !pip install -q scipy

Load the Embedding Model:¶

In [ ]:

from sentence_transformers import SentenceTransformer
model = SentenceTransformer("Alibaba-NLP/gte-base-en-v1.5", trust_remote_code=True)

from sentence_transformers import SentenceTransformer model = SentenceTransformer("Alibaba-NLP/gte-base-en-v1.5", trust_remote_code=True)

Fetch Text Content from Wikipedia:¶

In [ ]:

from wikipediaapi import Wikipedia
wiki = Wikipedia('RAGBot/0.0', 'en')
doc = wiki.page('Hayao_Miyazaki').text
paragraphs = doc.split('\n\n') # chunking

from wikipediaapi import Wikipedia wiki = Wikipedia('RAGBot/0.0', 'en') doc = wiki.page('Hayao_Miyazaki').text paragraphs = doc.split('\n\n') # chunking

In [ ]:

import textwrap

import textwrap

In [ ]:

for i, p in enumerate(paragraphs):
  wrapped_text = textwrap.fill(p, width=100)

  print("-----------------------------------------------------------------")
  print(wrapped_text)
  print("-----------------------------------------------------------------")

for i, p in enumerate(paragraphs): wrapped_text = textwrap.fill(p, width=100) print("-----------------------------------------------------------------") print(wrapped_text) print("-----------------------------------------------------------------")

Embed the Document:¶

In [ ]:

docs_embed = model.encode(paragraphs, normalize_embeddings=True)

docs_embed = model.encode(paragraphs, normalize_embeddings=True)

In [ ]:

docs_embed.shape

docs_embed.shape

In [ ]:

docs_embed[0]

docs_embed[0]

Embed the Query:¶

In [ ]:

query = "What was Studio Ghibli's first film?"
query_embed = model.encode(query, normalize_embeddings=True)

query = "What was Studio Ghibli's first film?" query_embed = model.encode(query, normalize_embeddings=True)

In [ ]:

query_embed.shape

query_embed.shape

Find the Closest Paragraphs to the Query:¶

In [ ]:

import numpy as np
similarities = np.dot(docs_embed, query_embed.T)

import numpy as np similarities = np.dot(docs_embed, query_embed.T)

In [ ]:

similarities.shape

similarities.shape

In [ ]:

similarities

similarities

In [ ]:

top_3_idx = np.argsort(similarities, axis=0)[-3:][::-1].tolist()

top_3_idx = np.argsort(similarities, axis=0)[-3:][::-1].tolist()

In [ ]:

top_3_idx

top_3_idx

In [ ]:

most_similar_documents = [paragraphs[idx] for idx in top_3_idx]

most_similar_documents = [paragraphs[idx] for idx in top_3_idx]

In [ ]:

CONTEXT = ""
for i, p in enumerate(most_similar_documents):
  wrapped_text = textwrap.fill(p, width=100)

  print("-----------------------------------------------------------------")
  print(wrapped_text)
  print("-----------------------------------------------------------------")
  CONTEXT += wrapped_text + "\n\n"

CONTEXT = "" for i, p in enumerate(most_similar_documents): wrapped_text = textwrap.fill(p, width=100) print("-----------------------------------------------------------------") print(wrapped_text) print("-----------------------------------------------------------------") CONTEXT += wrapped_text + "\n\n"

In [ ]:

query = "What was Studio Ghibli's first film?"

query = "What was Studio Ghibli's first film?"

In [ ]:

prompt = f"""
use the following CONTEXT to answer the QUESTION at the end.
If you don't know the answer, just say that you don't know, don't try to make up an answer.

CONTEXT: {CONTEXT}
QUESTION: {query}

"""

prompt = f""" use the following CONTEXT to answer the QUESTION at the end. If you don't know the answer, just say that you don't know, don't try to make up an answer. CONTEXT: {CONTEXT} QUESTION: {query} """

In [ ]:

!pip install -q openai

!pip install -q openai

In [ ]:

# prompt: write python code to make calls to openai api
from google.colab import userdata
userdata.get('openai')

import openai

# prompt: write python code to make calls to openai api from google.colab import userdata userdata.get('openai') import openai

In [ ]:

from openai import OpenAI
client = OpenAI(api_key=userdata.get('openai'))

from openai import OpenAI client = OpenAI(api_key=userdata.get('openai'))

In [ ]:

response = client.chat.completions.create(
  model="gpt-4o",
  messages=[
    {"role": "user", "content": prompt},
  ]
)

response = client.chat.completions.create( model="gpt-4o", messages=[ {"role": "user", "content": prompt}, ] )

In [ ]:

print(response.choices[0].message.content)

print(response.choices[0].message.content)

In [ ]: