bing chatbot

1 ActiveLoop Courses/1 LangChain & Vector Databases in Production/7 Making LLMs Interact with the World Using Tools/3 Recreating the Bing Chatbot/chatbot.py

@@ -0,0 +1,133 @@
+from langchain.prompts import  PromptTemplate
+from langchain.chains import LLMChain
+from langchain.llms import OpenAI
+
+llm = OpenAI(temperature=0)
+
+template = """You are an assistant that answers the following question correctly and honestly: {question}\n\n"""
+prompt_template = PromptTemplate(input_variables=["question"], template=template)
+
+question_chain = LLMChain(llm=llm, prompt=prompt_template, verbose=True)
+
+response = question_chain.run("what is the latest fast and furious movie?")
+print(response)
+
+print ("----- GETTING SEARCH RESULTS -----")
+
+
+
+
+from langchain.tools import Tool
+from langchain.utilities import GoogleSearchAPIWrapper
+
+search = GoogleSearchAPIWrapper()
+TOP_N_RESULTS = 10
+
+def top_n_results(query):
+    return search.results(query, TOP_N_RESULTS)
+
+tool = Tool(
+    name = "Google Search",
+    description="Search Google for recent results.",
+    func=top_n_results
+)
+
+query = "What is the latest fast and furious movie?"
+
+results = tool.run(query)
+
+for result in results:
+    print(result["title"])
+    print(result["link"])
+    print(result["snippet"])
+    print("-"*50)
+
+
+print ("------- DOWNLOAD AND PARSE CONTENTS -----")
+
+
+import newspaper
+
+pages_content = []
+
+for result in results:
+	try:
+		article = newspaper.Article(result["link"])
+		article.download()
+		article.parse()
+		if len(article.text) > 0:
+			pages_content.append({ "url": result["link"], "text": article.text })
+	except:
+		continue
+
+
+print ("----- PROCESS SEARCH RESULTS -----")
+
+
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain.docstore.document import Document
+
+text_splitter = RecursiveCharacterTextSplitter(chunk_size=4000, chunk_overlap=100)
+
+docs = []
+for d in pages_content:
+	chunks = text_splitter.split_text(d["text"])
+	for chunk in chunks:
+		new_doc = Document(page_content=chunk, metadata={ "source": d["url"] })
+		docs.append(new_doc)
+
+
+
+
+
+
+from langchain.embeddings import OpenAIEmbeddings
+
+embeddings = OpenAIEmbeddings(model="text-embedding-ada-002")
+
+docs_embeddings = embeddings.embed_documents([doc.page_content for doc in docs])
+query_embedding = embeddings.embed_query(query)
+
+
+
+import numpy as np
+from sklearn.metrics.pairwise import cosine_similarity
+
+def get_top_k_indices(list_of_doc_vectors, query_vector, top_k):
+  # convert the lists of vectors to numpy arrays
+  list_of_doc_vectors = np.array(list_of_doc_vectors)
+  query_vector = np.array(query_vector)
+
+  # compute cosine similarities
+  similarities = cosine_similarity(query_vector.reshape(1, -1), list_of_doc_vectors).flatten()
+
+  # sort the vectors based on cosine similarity
+  sorted_indices = np.argsort(similarities)[::-1]
+
+  # retrieve the top K indices from the sorted list
+  top_k_indices = sorted_indices[:top_k]
+
+  return top_k_indices
+
+top_k = 2
+best_indexes = get_top_k_indices(docs_embeddings, query_embedding, top_k)
+best_k_documents = [doc for i, doc in enumerate(docs) if i in best_indexes]
+
+
+print("---- CHAIN WITH SOURCE ---")
+
+from langchain.chains.qa_with_sources import load_qa_with_sources_chain
+from langchain.llms import OpenAI
+
+chain = load_qa_with_sources_chain(OpenAI(temperature=0), chain_type="stuff")
+
+response = chain({"input_documents": best_k_documents, "question": query}, return_only_outputs=True)
+
+response_text, response_sources = response["output_text"].split("SOURCES:")
+response_text = response_text.strip()
+response_sources = response_sources.strip()
+
+print(f"Answer: {response_text}")
+print(f"Sources: {response_sources}")
+
+

1 ActiveLoop Courses/1 LangChain & Vector Databases in Production/7 Making LLMs Interact with the World Using Tools/4 Integrating Multiple Tools for Web Based Question Answering

@@ -0,0 +1,11 @@
+from langchain import LLMChain, PromptTemplate
+from langchain.llms import OpenAI
+
+llm = OpenAI(temperature=0)
+
+template = """You are an assistant that answers the following question correctly and honestly: {question}\n\n"""
+prompt_template = PromptTemplate(input_variables=["question"], template=template)
+
+question_chain = LLMChain(llm=llm, prompt=prompt_template)
+
+question_chain.run("what is the latest fast and furious movie?")