Refactor imports and update HuggingFaceEndpoint configuration in app.py

app.py

@@ -1,40 +1,23 @@
 import os
-import torch
-from transformers import (
-  BitsAndBytesConfig,
-  pipeline
-)
 import streamlit as st
 from langchain_community.vectorstores import FAISS
 from langchain_community.embeddings import HuggingFaceEmbeddings
-from langchain_community.llms import HuggingFacePipeline
-from transformers import BitsAndBytesConfig
-from langchain.embeddings.huggingface import HuggingFaceEmbeddings
-from langchain_community.llms import HuggingFaceEndpoint
+
+from langchain_huggingface import HuggingFaceEndpoint
 
 from langchain.prompts import PromptTemplate
 from langchain.schema.runnable import RunnablePassthrough
 from langchain.chains import LLMChain
-import transformers
-from ctransformers import AutoModelForCausalLM, AutoTokenizer
 
-import transformers
-from transformers import pipeline
-from datasets import load_dataset
-
-import transformers
-token=st.secrets["HF_TOKEN"]
 from huggingface_hub import login
 login(token=st.secrets["HF_TOKEN"])
-# Load model directly
-from transformers import AutoTokenizer, AutoModelForCausalLM
-
-tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3")
-model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3")
 
 from langchain_community.document_loaders import TextLoader
 from langchain_text_splitters import CharacterTextSplitter
 from langchain_community.document_loaders import PyPDFLoader
+from langchain.chains import RetrievalQA
+from langchain.prompts import PromptTemplate
+from langchain.embeddings.huggingface import HuggingFaceEmbeddings
 
 # Montez Google Drive
 loader = PyPDFLoader("test-1.pdf")
@@ -53,21 +36,6 @@ retriever = db.as_retriever(
 )
 
 
-from langchain_community.llms import HuggingFacePipeline
-from langchain.prompts import PromptTemplate
-from langchain.embeddings.huggingface import HuggingFaceEmbeddings
-
-text_generation_pipeline = transformers.pipeline(
-    model=model,
-    tokenizer=tokenizer,
-    task="text-generation",
-
-    temperature=0.02,
-    repetition_penalty=1.1,
-    return_full_text=True,
-    max_new_tokens=512,
-)
-
 prompt_template = """
 ### [INST]
 Instruction: You are a Q&A assistant. Your goal is to answer questions as accurately as possible based on the instructions and context provided without using prior knowledge.You answer in FRENCH
@@ -84,7 +52,11 @@ Answer in french only
 
  """
 
-mistral_llm = HuggingFacePipeline(pipeline=text_generation_pipeline)
+repo_id = "mistralai/Mistral-7B-Instruct-v0.2"
+
+mistral_llm = HuggingFaceEndpoint(
+    repo_id=repo_id, max_length=128, temperature=0.5, huggingfacehub_api_token=st.secrets["HF_TOKEN"]
+)
 
 # Create prompt from prompt template
 prompt = PromptTemplate(
@@ -93,7 +65,7 @@ prompt = PromptTemplate(
 )
 
 # Create llm chain
-from langchain.chains import RetrievalQA
+llm_chain = LLMChain(llm=mistral_llm, prompt=prompt)
 
 
 retriever.search_kwargs = {'k':1}
@@ -111,7 +83,7 @@ st.title("Chatbot Interface")
 
 # Define function to handle user input and display chatbot response
 def chatbot_response(user_input):
-    response = qa.get_answer(user_input)
+    response = qa.run(user_input)
     return response
 
 # Streamlit components
@@ -124,4 +96,4 @@ if submit_button:
         bot_response = chatbot_response(user_input)
         st.text_area("Bot:", value=bot_response, height=200)
     else:
-        st.warning("Please enter a message.")
+        st.warning("Please enter a message.")

requirements.txt

@@ -1,5 +1,4 @@
-bitsandbytes
-peft
+langchain_huggingface
 sentence_transformers
 scipy
 langchain