rag.py

import os
from langchain.chat_models import ChatOpenAI
from langchain.schema import (
    SystemMessage,
    HumanMessage,
    AIMessage
)
from datasets import load_dataset
from pinecone import Pinecone
from pinecone import ServerlessSpec
import time
from langchain_openai import OpenAIEmbeddings
from tqdm.auto import tqdm  


dataset = load_dataset(
    "jamescalam/llama-2-arxiv-papers-chunked",
    split="train"
)

os.environ["OPENAI_API_KEY"] = os.getenv("OPENAI_API_KEY") 

chat = ChatOpenAI(
    openai_api_key = os.environ["OPENAI_API_KEY"],
    model='gpt-3.5-turbo'

)


messages = [
    SystemMessage(content="You are a helpful assistant."),
    HumanMessage(content="Hi AI, how are you today?"),
    AIMessage(content="I'm great thank you. How can I help you?"),
    HumanMessage(content="I'd like to understand string theory.")
]

res = chat(messages)

# add latest AI response to messages
messages.append(res)

# connect to pinecone
api_key = os.getenv('PINECONE_API_KEY')

# configure client
pc = Pinecone(api_key=api_key)

# connect to serverless
spec = ServerlessSpec(
    cloud="aws", region="us-east-1"
)

# initialize index
index_name = 'llama-2-rag'
existing_indexes = [
    index_info["name"] for index_info in pc.list_indexes()
]

# check if index already exists (it shouldn't if this is first time)
if index_name not in existing_indexes:
    # if does not exist, create index
    pc.create_index(
        index_name,
        dimension=1536,  # dimensionality of ada 002
        metric='dotproduct',
        spec=spec
    )
    # wait for index to be initialized
    while not pc.describe_index(index_name).status['ready']:
        time.sleep(1)

# connect to index
index = pc.Index(index_name)
time.sleep(1)
# view index stats
index.describe_index_stats()

# create vector embeddings of our index
embed_model = OpenAIEmbeddings(model="text-embedding-ada-002")

# iterate over dataset
data = dataset.to_pandas()
batch_size = 100

for i in tqdm(range(0, len(data), batch_size)):
    i_end = min(len(data), i+batch_size)
    # get batch of data
    batch = data.iloc[i:i_end]
    # generate unique ids for each chunk
    ids = [f"{x['doi']}-{x['chunk-id']}" for i, x in batch.iterrows()]
    # get text to embed
    texts = [x['chunk'] for _, x in batch.iterrows()]
    # embed text
    embeds = embed_model.embed_documents(texts)
    # get metadata to store in Pinecone
    metadata = [
        {'text': x['chunk'],
         'source': x['source'],
         'title': x['title']} for i, x in batch.iterrows()
    ]
    # add to Pinecone
    index.upsert(vectors=zip(ids, embeds, metadata))

index.describe_index_stats()

#### Retrival Augmented Generation
#from langchain_pinecone import PineconeVectoreStore
from langchain.vectorstores import Pinecone

# the metadata field that contains our text
text_field = "text"  

# initialize the vector store object
vectorstore = Pinecone(
    index, embed_model.embed_query, text_field
)

query = "What is so special about Llama 2?"

vectorstore.similarity_search(query, k=3)

# connect the output from vectorstore to chat 
def augment_prompt(query: str):
    # get top 3 results from knowledge base
    results = vectorstore.similarity_search(query, k=3)
    # get the text from the results
    source_knowledge = "\n".join([x.page_content for x in results])
    # feed into an augmented prompt
    augmented_prompt = f"""Using the contexts below, answer the query.

    Contexts:
    {source_knowledge}

    Query: {query}"""
    return augmented_prompt

# create a new user prompt
prompt = HumanMessage(
    content=augment_prompt(query)
)
# add to messages
messages.append(prompt)

res = chat(messages)

print(res.content)