Update app.py

app.py

@@ -1,273 +1,47 @@
-# import os
-# import time
-# from fastapi import FastAPI,Request
-# from fastapi.responses import HTMLResponse
-# from fastapi.staticfiles import StaticFiles
-# from llama_index.core import StorageContext, load_index_from_storage, VectorStoreIndex, SimpleDirectoryReader, ChatPromptTemplate, Settings
-# from llama_index.embeddings.huggingface import HuggingFaceEmbedding
-# from pydantic import BaseModel
-# from fastapi.responses import JSONResponse
-# import uuid  # for generating unique IDs
-# import datetime
-# from fastapi.middleware.cors import CORSMiddleware
-# from fastapi.templating import Jinja2Templates
-# from huggingface_hub import InferenceClient
-# import json
-# import re
-# from gradio_client import Client
-# from simple_salesforce import Salesforce, SalesforceLogin
-# from llama_index.llms.huggingface import HuggingFaceLLM
-
-
-# # Define Pydantic model for incoming request body
-# class MessageRequest(BaseModel):
-#     message: str
-# # repo_id = "meta-llama/Meta-Llama-3-8B-Instruct"
-# # llm_client = InferenceClient(
-# #     model=repo_id,
-# #     token=os.getenv("HF_TOKEN"),
-# # )
-
-
-# os.environ["HF_TOKEN"] = os.getenv("HF_TOKEN")
-# username = os.getenv("username")
-# password = os.getenv("password")
-# security_token = os.getenv("security_token")
-# domain =  os.getenv("domain")# Using sandbox environment
-# session_id, sf_instance = SalesforceLogin(username=username, password=password, security_token=security_token, domain=domain)
-
-#     # Create Salesforce object
-# sf = Salesforce(instance=sf_instance, session_id=session_id)
-
-# app = FastAPI()
-
-
-# @app.middleware("http")
-# async def add_security_headers(request: Request, call_next):
-#     response = await call_next(request)
-#     response.headers["Content-Security-Policy"] = "frame-ancestors *; frame-src *; object-src *;"
-#     response.headers["X-Frame-Options"] = "ALLOWALL"
-#     return response
-
-
-# # Allow CORS requests from any domain
-# app.add_middleware(
-#     CORSMiddleware,
-#     allow_origins=["*"],
-#     allow_credentials=True,
-#     allow_methods=["*"],
-#     allow_headers=["*"],
-# )
-
-
-
-
-# @app.get("/favicon.ico")
-# async def favicon():
-#     return HTMLResponse("")  # or serve a real favicon if you have one
-
-
-# app.mount("/static", StaticFiles(directory="static"), name="static")
-
-# templates = Jinja2Templates(directory="static")
-# # Configure Llama index settings
-# # Settings.llm = HuggingFaceInferenceAPI(
-# #     model_name="meta-llama/Meta-Llama-3-8B-Instruct",
-# #     tokenizer_name="meta-llama/Meta-Llama-3-8B-Instruct",
-# #     context_window=3000,
-# #     token=os.getenv("HF_TOKEN"),
-# #     max_new_tokens=512,
-# #     generate_kwargs={"temperature": 0.1},
-# # )
-# # Configure Llama index settings
-# Settings.llm = HuggingFaceLLM(
-#     model_name="google/flan-t5-small",
-#     tokenizer_name="google/flan-t5-small",
-#     context_window=512,  # flan-t5-small has a max context window of 512 tokens
-#     max_new_tokens=256,
-#     generate_kwargs={"temperature": 0.1, "do_sample": True},
-#     device_map="auto"  # Automatically use GPU if available, else CPU
-# )
-# Settings.embed_model = HuggingFaceEmbedding(
-#     model_name="BAAI/bge-small-en-v1.5"
-# )
-
-# PERSIST_DIR = "db"
-# PDF_DIRECTORY = 'data'
-
-# # Ensure directories exist
-# os.makedirs(PDF_DIRECTORY, exist_ok=True)
-# os.makedirs(PERSIST_DIR, exist_ok=True)
-# chat_history = []
-# current_chat_history = []
-# def data_ingestion_from_directory():
-#     documents = SimpleDirectoryReader(PDF_DIRECTORY).load_data()
-#     storage_context = StorageContext.from_defaults()
-#     index = VectorStoreIndex.from_documents(documents)
-#     index.storage_context.persist(persist_dir=PERSIST_DIR)
-
-# def initialize():
-#     start_time = time.time()
-#     data_ingestion_from_directory()  # Process PDF ingestion at startup
-#     print(f"Data ingestion time: {time.time() - start_time} seconds")
-# def split_name(full_name):
-#     # Split the name by spaces
-#     words = full_name.strip().split()
-    
-#     # Logic for determining first name and last name
-#     if len(words) == 1:
-#         first_name = ''
-#         last_name = words[0]
-#     elif len(words) == 2:
-#         first_name = words[0]
-#         last_name = words[1]
-#     else:
-#         first_name = words[0]
-#         last_name = ' '.join(words[1:])
-    
-#     return first_name, last_name
-
-# initialize()  # Run initialization tasks
-
-
-# def handle_query(query):
-#     chat_text_qa_msgs = [
-#         (
-#             "user",
-#             """
-#             You are the Clara Redfernstech chatbot. Your goal is to provide accurate, professional, and helpful answers to user queries based on the company's data. Always ensure your responses are clear and concise. Give response within 10-15 words only       
-#             {context_str}
-#             Question:
-#             {query_str}
-#             """
-#         )
-#     ]
-#     text_qa_template = ChatPromptTemplate.from_messages(chat_text_qa_msgs)
-    
-#     storage_context = StorageContext.from_defaults(persist_dir=PERSIST_DIR)
-#     index = load_index_from_storage(storage_context)
-#     context_str = ""
-#     for past_query, response in reversed(current_chat_history):
-#         if past_query.strip():
-#             context_str += f"User asked: '{past_query}'\nBot answered: '{response}'\n"
-
-    
-#     query_engine = index.as_query_engine(text_qa_template=text_qa_template, context_str=context_str)
-#     answer = query_engine.query(query)
-
-#     if hasattr(answer, 'response'):
-#         response=answer.response
-#     elif isinstance(answer, dict) and 'response' in answer:
-#         response =answer['response']
-#     else:
-#         response ="Sorry, I couldn't find an answer."
-#     current_chat_history.append((query, response))
-#     return response
-# @app.get("/ch/{id}", response_class=HTMLResponse)
-# async def load_chat(request: Request, id: str):
-#     return templates.TemplateResponse("index.html", {"request": request, "user_id": id})
-# # Route to save chat history
-# @app.post("/hist/")
-# async def save_chat_history(history: dict):
-#     # Check if 'userId' is present in the incoming dictionary
-#     user_id = history.get('userId')
-#     print(user_id)
-
-#     # Ensure user_id is defined before proceeding
-#     if user_id is None:
-#         return {"error": "userId is required"}, 400
-
-#     # Construct the chat history string
-#     hist = ''.join([f"'{entry['sender']}: {entry['message']}'\n" for entry in history['history']])
-#     hist = "You are a Redfernstech summarize model. Your aim is to use this conversation to identify user interests solely based on that conversation: " + hist
-#     print(hist)
-
-#     # Get the summarized result from the client model
-#     result = hist
-
-#     try:
-#         sf.Lead.update(user_id, {'Description': result})
-#     except Exception as e:
-#         return {"error": f"Failed to update lead: {str(e)}"}, 500
-    
-#     return {"summary": result, "message": "Chat history saved"}
-# @app.post("/webhook")
-# async def receive_form_data(request: Request):
-#     form_data = await request.json()
-#     # Log in to Salesforce
-#     first_name, last_name = split_name(form_data['name'])
-#     data = {
-#     'FirstName': first_name,
-#     'LastName': last_name,
-#     'Description': 'hii',  # Static description
-#     'Company': form_data['company'],  # Assuming company is available in form_data
-#     'Phone': form_data['phone'].strip(),  # Phone from form data
-#     'Email': form_data['email'],  # Email from form data
-#     }
-#     a=sf.Lead.create(data)
-#     # Generate a unique ID (for tracking user)
-#     unique_id = a['id']
-    
-#     # Here you can do something with form_data like saving it to a database
-#     print("Received form data:", form_data)
-    
-#     # Send back the unique id to the frontend
-#     return JSONResponse({"id": unique_id})
-
-# @app.post("/chat/")
-# async def chat(request: MessageRequest):
-#     message = request.message  # Access the message from the request body
-#     response = handle_query(message)  # Process the message
-#     message_data = {
-#         "sender": "User",
-#         "message": message,
-#         "response": response,
-#         "timestamp": datetime.datetime.now().isoformat()
-#     }
-#     chat_history.append(message_data)
-#     return {"response": response}
-# @app.get("/")
-# def read_root():
-#     return {"message": "Welcome to the API"}
-
 import os
 import time
-import logging
-from fastapi import FastAPI, Request
+from fastapi import FastAPI,Request
 from fastapi.responses import HTMLResponse
 from fastapi.staticfiles import StaticFiles
-from llama_index.core import StorageContext, load_index_from_storage, VectorStoreIndex, SimpleDirectoryReader, PromptTemplate, Settings
-from llama_index.llms.huggingface import HuggingFaceLLM
+from llama_index.core import StorageContext, load_index_from_storage, VectorStoreIndex, SimpleDirectoryReader, ChatPromptTemplate, Settings
 from llama_index.embeddings.huggingface import HuggingFaceEmbedding
 from pydantic import BaseModel
 from fastapi.responses import JSONResponse
+import uuid  # for generating unique IDs
 import datetime
 from fastapi.middleware.cors import CORSMiddleware
 from fastapi.templating import Jinja2Templates
+from huggingface_hub import InferenceClient
+import json
+import re
+from gradio_client import Client
 from simple_salesforce import Salesforce, SalesforceLogin
-from transformers import AutoModelForSeq2SeqLM, AutoTokenizer
+from llama_index.llms.huggingface import HuggingFaceLLM
 
-# Set up logging
-logging.basicConfig(level=logging.INFO)
-logger = logging.getLogger(__name__)
 
 # Define Pydantic model for incoming request body
 class MessageRequest(BaseModel):
     message: str
+repo_id = "meta-llama/Meta-Llama-3-8B-Instruct"
+llm_client = InferenceClient(
+    model=repo_id,
+    token=os.getenv("HF_TOKEN"),
+)
+
 
-# Environment variables for Salesforce
-os.environ["HF_TOKEN"] = os.getenv("HF_TOKEN", "")  # Optional: Needed for model download if gated
+os.environ["HF_TOKEN"] = os.getenv("HF_TOKEN")
 username = os.getenv("username")
 password = os.getenv("password")
 security_token = os.getenv("security_token")
-domain = os.getenv("domain", "test")  # Default to sandbox environment
+domain =  os.getenv("domain")# Using sandbox environment
 session_id, sf_instance = SalesforceLogin(username=username, password=password, security_token=security_token, domain=domain)
 
-# Create Salesforce object
+    # Create Salesforce object
 sf = Salesforce(instance=sf_instance, session_id=session_id)
 
 app = FastAPI()
 
+
 @app.middleware("http")
 async def add_security_headers(request: Request, call_next):
     response = await call_next(request)
@@ -275,6 +49,7 @@ async def add_security_headers(request: Request, call_next):
     response.headers["X-Frame-Options"] = "ALLOWALL"
     return response
 
+
 # Allow CORS requests from any domain
 app.add_middleware(
     CORSMiddleware,
@@ -284,162 +59,157 @@ app.add_middleware(
     allow_headers=["*"],
 )
 
+
+
+
 @app.get("/favicon.ico")
 async def favicon():
     return HTMLResponse("")  # or serve a real favicon if you have one
 
+
 app.mount("/static", StaticFiles(directory="static"), name="static")
 
 templates = Jinja2Templates(directory="static")
-
-# Configure Llama index settings
-tokenizer = AutoTokenizer.from_pretrained("google/flan-t5-small")
-Settings.llm = HuggingFaceLLM(
-    model_name="google/flan-t5-small",
-    tokenizer_name="google/flan-t5-small",
-    context_window=512,  # flan-t5-small has a max context window of 512 tokens
-    max_new_tokens=256,
-    generate_kwargs={"temperature": 0.3, "do_sample": True},  # Increased temperature for better responses
-    model=AutoModelForSeq2SeqLM.from_pretrained("google/flan-t5-small"),
-    tokenizer=tokenizer,
-    device_map="auto"  # Automatically use GPU if available, else CPU
+Configure Llama index settings
+Settings.llm = HuggingFaceInferenceAPI(
+    model_name="meta-llama/Meta-Llama-3-8B-Instruct",
+    tokenizer_name="meta-llama/Meta-Llama-3-8B-Instruct",
+    context_window=3000,
+    token=os.getenv("HF_TOKEN"),
+    max_new_tokens=512,
+    generate_kwargs={"temperature": 0.1},
 )
+# Configure Llama index settings
+# Settings.llm = HuggingFaceLLM(
+#     model_name="google/flan-t5-small",
+#     tokenizer_name="google/flan-t5-small",
+#     context_window=512,  # flan-t5-small has a max context window of 512 tokens
+#     max_new_tokens=256,
+#     generate_kwargs={"temperature": 0.1, "do_sample": True},
+#     device_map="auto"  # Automatically use GPU if available, else CPU
+# )
 Settings.embed_model = HuggingFaceEmbedding(
     model_name="BAAI/bge-small-en-v1.5"
 )
 
 PERSIST_DIR = "db"
-PDF_DIRECTORY = "data"
+PDF_DIRECTORY = 'data'
 
 # Ensure directories exist
 os.makedirs(PDF_DIRECTORY, exist_ok=True)
 os.makedirs(PERSIST_DIR, exist_ok=True)
 chat_history = []
-current_chat_history = [
-    ("What is Redfernstech?", "Redfernstech is a technology company specializing in AI solutions."),
-    ("What services does Redfernstech offer?", "Redfernstech offers AI-powered chatbots, data analytics, and automation tools.")
-]
-
+current_chat_history = []
 def data_ingestion_from_directory():
-    try:
-        documents = SimpleDirectoryReader(PDF_DIRECTORY).load_data()
-        logger.info(f"Loaded {len(documents)} documents from {PDF_DIRECTORY}")
-        if not documents:
-            logger.warning(f"No documents found in {PDF_DIRECTORY}. Ensure PDF files are present.")
-        storage_context = StorageContext.from_defaults()
-        index = VectorStoreIndex.from_documents(documents)
-        index.storage_context.persist(persist_dir=PERSIST_DIR)
-        logger.info(f"Index persisted to {PERSIST_DIR}")
-    except Exception as e:
-        logger.error(f"Error during data ingestion: {str(e)}")
-        raise
+    documents = SimpleDirectoryReader(PDF_DIRECTORY).load_data()
+    storage_context = StorageContext.from_defaults()
+    index = VectorStoreIndex.from_documents(documents)
+    index.storage_context.persist(persist_dir=PERSIST_DIR)
 
 def initialize():
     start_time = time.time()
     data_ingestion_from_directory()  # Process PDF ingestion at startup
-    logger.info(f"Data ingestion time: {time.time() - start_time} seconds")
-
+    print(f"Data ingestion time: {time.time() - start_time} seconds")
 def split_name(full_name):
     # Split the name by spaces
     words = full_name.strip().split()
     
     # Logic for determining first name and last name
     if len(words) == 1:
-        first_name = ""
+        first_name = ''
         last_name = words[0]
     elif len(words) == 2:
         first_name = words[0]
         last_name = words[1]
     else:
         first_name = words[0]
-        last_name = " ".join(words[1:])
+        last_name = ' '.join(words[1:])
     
     return first_name, last_name
 
 initialize()  # Run initialization tasks
 
+
 def handle_query(query):
-    # Custom prompt template for flan-t5-small
-    text_qa_template = PromptTemplate(
-        """
-        You are Clara, a Redfernstech chatbot. Answer the question in 10-15 words based on the provided context.
-        Context: {context_str}
-        Question: {query_str}
-        Answer:
-        """
-    )
+    chat_text_qa_msgs = [
+        (
+            "user",
+            """
+            You are the Clara Redfernstech chatbot. Your goal is to provide accurate, professional, and helpful answers to user queries based on the company's data. Always ensure your responses are clear and concise. Give response within 10-15 words only       
+            {context_str}
+            Question:
+            {query_str}
+            """
+        )
+    ]
+    text_qa_template = ChatPromptTemplate.from_messages(chat_text_qa_msgs)
     
     storage_context = StorageContext.from_defaults(persist_dir=PERSIST_DIR)
     index = load_index_from_storage(storage_context)
+    context_str = ""
+    for past_query, response in reversed(current_chat_history):
+        if past_query.strip():
+            context_str += f"User asked: '{past_query}'\nBot answered: '{response}'\n"
+
     
-    # Build context from chat history
-    context_str = "\n".join([f"User asked: '{past_query}'\nBot answered: '{response}'" 
-                             for past_query, response in reversed(current_chat_history)])
-    
-    logger.info(f"Query: {query}")
-    logger.info(f"Context: {context_str}")
-    
-    query_engine = index.as_query_engine(text_qa_template=text_qa_template)
+    query_engine = index.as_query_engine(text_qa_template=text_qa_template, context_str=context_str)
     answer = query_engine.query(query)
-    
-    logger.info(f"Raw query engine output: {answer}")
-    
-    if hasattr(answer, "response") and answer.response:
-        response = answer.response.strip()
-    elif isinstance(answer, dict) and "response" in answer and answer["response"]:
-        response = answer["response"].strip()
+
+    if hasattr(answer, 'response'):
+        response=answer.response
+    elif isinstance(answer, dict) and 'response' in answer:
+        response =answer['response']
     else:
-        response = "Sorry, I couldn't find an answer."
-    
-    logger.info(f"Processed response: {response}")
+        response ="Sorry, I couldn't find an answer."
     current_chat_history.append((query, response))
     return response
-
 @app.get("/ch/{id}", response_class=HTMLResponse)
 async def load_chat(request: Request, id: str):
     return templates.TemplateResponse("index.html", {"request": request, "user_id": id})
-
+# Route to save chat history
 @app.post("/hist/")
 async def save_chat_history(history: dict):
     # Check if 'userId' is present in the incoming dictionary
-    user_id = history.get("userId")
-    logger.info(f"Received userId: {user_id}")
+    user_id = history.get('userId')
+    print(user_id)
 
     # Ensure user_id is defined before proceeding
     if user_id is None:
         return {"error": "userId is required"}, 400
 
     # Construct the chat history string
-    hist = "".join([f"'{entry['sender']}: {entry['message']}'\n" for entry in history["history"]])
+    hist = ''.join([f"'{entry['sender']}: {entry['message']}'\n" for entry in history['history']])
     hist = "You are a Redfernstech summarize model. Your aim is to use this conversation to identify user interests solely based on that conversation: " + hist
-    logger.info(f"Chat history: {hist}")
+    print(hist)
+
+    # Get the summarized result from the client model
+    result = hist
 
     try:
-        sf.Lead.update(user_id, {"Description": hist})
+        sf.Lead.update(user_id, {'Description': result})
     except Exception as e:
-        logger.error(f"Failed to update lead: {str(e)}")
         return {"error": f"Failed to update lead: {str(e)}"}, 500
     
-    return {"summary": hist, "message": "Chat history saved"}
-
+    return {"summary": result, "message": "Chat history saved"}
 @app.post("/webhook")
 async def receive_form_data(request: Request):
     form_data = await request.json()
     # Log in to Salesforce
-    first_name, last_name = split_name(form_data["name"])
+    first_name, last_name = split_name(form_data['name'])
     data = {
-        "FirstName": first_name,
-        "LastName": last_name,
-        "Description": f"New lead created via webhook: {form_data['name']}",
-        "Company": form_data.get("company", ""),  # Assuming company might be optional
-        "Phone": form_data["phone"].strip(),  # Phone from form data
-        "Email": form_data["email"],  # Email from form data
+    'FirstName': first_name,
+    'LastName': last_name,
+    'Description': 'hii',  # Static description
+    'Company': form_data['company'],  # Assuming company is available in form_data
+    'Phone': form_data['phone'].strip(),  # Phone from form data
+    'Email': form_data['email'],  # Email from form data
     }
-    a = sf.Lead.create(data)
+    a=sf.Lead.create(data)
     # Generate a unique ID (for tracking user)
-    unique_id = a["id"]
+    unique_id = a['id']
     
-    logger.info(f"Received form data: {form_data}")
+    # Here you can do something with form_data like saving it to a database
+    print("Received form data:", form_data)
     
     # Send back the unique id to the frontend
     return JSONResponse({"id": unique_id})
@@ -447,7 +217,6 @@ async def receive_form_data(request: Request):
 @app.post("/chat/")
 async def chat(request: MessageRequest):
     message = request.message  # Access the message from the request body
-    logger.info(f"Received chat message: {message}")
     response = handle_query(message)  # Process the message
     message_data = {
         "sender": "User",
@@ -456,9 +225,7 @@ async def chat(request: MessageRequest):
         "timestamp": datetime.datetime.now().isoformat()
     }
     chat_history.append(message_data)
-    logger.info(f"Chat response: {response}")
     return {"response": response}
-
 @app.get("/")
 def read_root():
     return {"message": "Welcome to the API"}