app.py

#####################################
##   
#####################################

from langchain_community.llms import HuggingFaceHub

#from langchain_community.llms import HuggingFaceHub

llm_zephyr_7b_beta = HuggingFaceHub(
    repo_id="HuggingFaceH4/zephyr-7b-beta",
    task="text-generation",
    model_kwargs={
        "max_new_tokens": 512,
        "top_k": 30,
        "temperature": 0.1,
        "repetition_penalty": 1.03,
    },
)

import os
from crewai import Agent, Task, Crew, Process
from crewai_tools.tools import SerperDevTool

#from crewai_tools import SerperDevTool

search_tool = SerperDevTool()

# Define your agents with roles and goals
researcher = Agent(
  role='Senior Research Analyst',
  goal='Uncover cutting-edge developments in AI and data science',
  backstory="""You work at a leading tech think tank.
  Your expertise lies in identifying emerging trends.
  You have a knack for dissecting complex data and presenting actionable insights.""",
  verbose=True,
  allow_delegation=False,
  tools=[search_tool],
  llm=llm_zephyr_7b_beta
  # You can pass an optional llm attribute specifying what mode you wanna use.
  # It can be a local model through Ollama / LM Studio or a remote
  # model like OpenAI, Mistral, Antrophic or others (https://docs.crewai.com/how-to/LLM-Connections/)
  #
  # import os
  # os.environ['OPENAI_MODEL_NAME'] = 'gpt-3.5-turbo'
  #
  # OR
  #
  # from langchain_openai import ChatOpenAI
  # llm=ChatOpenAI(model_name="gpt-3.5", temperature=0.7)
)

writer = Agent(
  role='Tech Content Strategist',
  goal='Craft compelling content on tech advancements',
  backstory="""You are a renowned Content Strategist, known for your insightful and engaging articles.
  You transform complex concepts into compelling narratives.""",
  verbose=True,
  allow_delegation=True,
  llm=llm_zephyr_7b_beta
)

# Create tasks for your agents
task1 = Task(
  description="""Conduct a comprehensive analysis of the latest advancements in AI in 2024.
  Identify key trends, breakthrough technologies, and potential industry impacts.""",
  expected_output="Full analysis report in bullet points",
  agent=researcher
)

task2 = Task(
  description="""Using the insights provided, develop an engaging blog
  post that highlights the most significant AI advancements.
  Your post should be informative yet accessible, catering to a tech-savvy audience.
  Make it sound cool, avoid complex words so it doesn't sound like AI.""",
  expected_output="Full blog post of at least 4 paragraphs",
  agent=writer
)

# Instantiate your crew with a sequential process
crew = Crew(
  agents=[researcher, writer],
  tasks=[task1, task2],
  verbose=2, # You can set it to 1 or 2 to different logging levels
)

# Get your crew to work!
#result = crew.kickoff()

#print("######################")
#print(result)

##################
###### other models:
# "Trelis/Llama-2-7b-chat-hf-sharded-bf16"
# "bn22/Mistral-7B-Instruct-v0.1-sharded"
# "HuggingFaceH4/zephyr-7b-beta"

# function for loading 4-bit quantized model
def load_model( ):

    model =  HuggingFaceHub(
        repo_id="mistralai/Mixtral-8x7B-Instruct-v0.1",
        model_kwargs={"max_length": 1048, "temperature":0.2, "max_new_tokens":256, "top_p":0.95, "repetition_penalty":1.0},
    )
    
    return model

##################################################
## vs chat
##################################################
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig, pipeline

from langchain_core.messages import AIMessage, HumanMessage
from langchain_community.document_loaders import WebBaseLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_community.vectorstores import Chroma

from langchain_community.embeddings import HuggingFaceBgeEmbeddings
from langchain.vectorstores.faiss import FAISS


from dotenv import load_dotenv
from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
from langchain.chains import create_history_aware_retriever, create_retrieval_chain
from langchain.chains.combine_documents import create_stuff_documents_chain


load_dotenv()

def get_vectorstore():
    ''' 
        FAISS
        A FAISS vector store containing the embeddings of the text chunks.
   '''
    model = "BAAI/bge-base-en-v1.5"
    encode_kwargs = {
        "normalize_embeddings": True
    }  # set True to compute cosine similarity
    embeddings = HuggingFaceBgeEmbeddings(
        model_name=model, encode_kwargs=encode_kwargs, model_kwargs={"device": "cpu"}
    )
    # load from disk
    vector_store = Chroma(persist_directory="/home/user/.cache/chroma_db", embedding_function=embeddings)
    return vector_store

def get_vectorstore_from_url(url):
    # get the text in document form
    loader = WebBaseLoader(url)
    document = loader.load()
    
    # split the document into chunks
    text_splitter = RecursiveCharacterTextSplitter()
    document_chunks = text_splitter.split_documents(document)
    ####### 
    ''' 
        FAISS
        A FAISS vector store containing the embeddings of the text chunks.
   '''
    model = "BAAI/bge-base-en-v1.5"
    encode_kwargs = {
        "normalize_embeddings": True
    }  # set True to compute cosine similarity
    embeddings = HuggingFaceBgeEmbeddings(
        model_name=model, encode_kwargs=encode_kwargs, model_kwargs={"device": "cpu"}
    )
    # load from disk
    #vector_store = Chroma(persist_directory="/home/user/.cache/chroma_db", embedding_function=embeddings)
 
    #vectorstore = FAISS.from_texts(texts=text_chunks, embedding=embeddings)
    vector_store = Chroma.from_documents(document_chunks, embeddings, persist_directory="/home/user/.cache/chroma_db")
    
    all_documents = vector_store.get()['documents']
    total_records = len(all_documents)
    print("Total records in the collection: ", total_records)

    return vector_store

def get_context_retriever_chain(vector_store): 

    llm = load_model( )
    
    retriever = vector_store.as_retriever()
    
    prompt = ChatPromptTemplate.from_messages([
      MessagesPlaceholder(variable_name="chat_history"),
      ("user", "{input}"),
      ("user", "Given the above conversation, generate a search query to look up in order to get information relevant to the conversation")
    ])
    
    retriever_chain = create_history_aware_retriever(llm, retriever, prompt)
    
    return retriever_chain
    
def get_conversational_rag_chain(retriever_chain): 
    
    llm = load_model( )
    
    prompt = ChatPromptTemplate.from_messages([
      ("system", "Du bist eine freundlicher Mitarbeiterin Namens Susie und arbeitest in einenm Call Center. Du beantwortest basierend auf dem Context. Benutze nur den Inhalt des Context. Füge wenn möglich die Quelle hinzu. Antworte mit: Ich bin mir nicht sicher. Wenn die Antwort nicht aus dem Context hervorgeht. Antworte auf Deutsch, bitte? CONTEXT:\n\n{context}"),
      MessagesPlaceholder(variable_name="chat_history"),
      ("user", "{input}"),
    ])
    
    stuff_documents_chain = create_stuff_documents_chain(llm,prompt)
    
    return create_retrieval_chain(retriever_chain, stuff_documents_chain)


###################

###################
import gradio as gr


chat_history = []     # Set your chat history here

# Define your function here
def get_response(user_input):

    vs = get_vectorstore()
    chat_history =[]
    retriever_chain = get_context_retriever_chain(vs)
    conversation_rag_chain = get_conversational_rag_chain(retriever_chain)
    
    response = conversation_rag_chain.invoke({
        "chat_history": chat_history,
        "input": user_input
    })
    #print("get_response " +response)
    res = response['answer']
    parts = res.split(" Assistant: ")
    last_part = parts[-1]
    return last_part


###############
#####
#####
#####
####
from fastapi import FastAPI
from fastapi.middleware.cors import CORSMiddleware

app = FastAPI()

# middlewares to allow cross orgin communications
app.add_middleware(
    CORSMiddleware,
    allow_origins=['*'], 
    allow_credentials=True, 
    allow_methods=['*'], 
    allow_headers=['*'],
)


@app.post("/generate/")
def generate(user_input, history=[]):
    print("----yuhu -----")   
    return get_response(user_input, history)
##################

def history_to_dialog_format(chat_history: list[str]):
    dialog = []
    if len(chat_history) > 0:
        for idx, message in enumerate(chat_history[0]):
            role = "user" if idx % 2 == 0 else "assistant"
            dialog.append({
                "role": role,
                "content": message,
            })
    return dialog

def get_response(message, history):
    dialog = history_to_dialog_format(history)
    dialog.append({"role": "user", "content": message})

      # Define the prompt as a ChatPromptValue object
    #user_input = ChatPromptValue(user_input)
    
    # Convert the prompt to a tensor
    #input_ids = user_input.tensor
    

    #vs = get_vectorstore_from_url(user_url, all_domain)
    vs = get_vectorstore()
  
    history =[]
    retriever_chain = get_context_retriever_chain(vs)
    conversation_rag_chain = get_conversational_rag_chain(retriever_chain)
    
    response = conversation_rag_chain.invoke({
        "chat_history": history,
        "input": message  + " Assistant: ",
        "chat_message": message + " Assistant: "
    })
    #print("get_response " +response)
    res = response['answer']
    parts = res.split(" Assistant: ")
    last_part = parts[-1]
    return last_part#[-1]['generation']['content']    




 

######

########
import requests
from bs4 import BeautifulSoup
from urllib.parse import urlparse, urljoin

def get_links_from_page(url, visited_urls, domain_links):
    if url in visited_urls:
        return

    if len(visited_urls) > 25:
        return

    visited_urls.add(url)
    print(url)
    response = requests.get(url)

    if response.status_code == 200:
        soup = BeautifulSoup(response.content, 'html.parser')
        base_url = urlparse(url).scheme + '://' + urlparse(url).netloc
        links = soup.find_all('a', href=True)

        for link in links:
            href = link.get('href')
            absolute_url = urljoin(base_url, href)
            parsed_url = urlparse(absolute_url)

            if parsed_url.netloc == urlparse(url).netloc:
                domain_links.add(absolute_url)
                get_links_from_page(absolute_url, visited_urls, domain_links)

    else:
        print(f"Failed to retrieve content from {url}. Status code: {response.status_code}")

def get_all_links_from_domain(domain_url):
    visited_urls = set()
    domain_links = set()
    get_links_from_page(domain_url, visited_urls, domain_links)
    return domain_links





def simple(text:str):
  return text +" hhhmmm "

fe_app = gr.ChatInterface(
    fn=get_response,
    #fn=simple,
   # inputs=["text"],
   # outputs="text",
    title="Chat with Websites",
    description="Schreibe hier deine Frage rein...",
    #allow_flagging=False
    retry_btn=None,
    undo_btn=None,
    clear_btn=None 
)

#fe_app.launch(debug=True, share=True) 



# load the model asynchronously on startup and save it into memory 
@app.on_event("startup")
async def startup():
    # Get your crew to work!
    result = crew.kickoff()

    print("######################")
    print(result)
    #domain_url = 'https://globl.contact/'
    #links = get_all_links_from_domain(domain_url)
    #print("Links from the domain:", links)
    
    #########
    # Assuming visited_urls is a list of URLs
    #for url in links:
    #    vs = get_vectorstore_from_url(url)
    #load_model()