Update app.py

app.py

@@ -1,7 +1,7 @@
 import streamlit as st
 from PyPDF2 import PdfReader
 from langchain.text_splitter import RecursiveCharacterTextSplitter
-from langchain_community.embeddings import HuggingFaceEmbeddings  # Updated import
+from langchain_community.embeddings import HuggingFaceEmbeddings
 from langchain.vectorstores import FAISS
 from langchain.chains import RetrievalQA
 from langchain.llms import HuggingFacePipeline
@@ -9,10 +9,18 @@ import torch
 from transformers import pipeline
 from langdetect import detect
 
-# Load a smaller LLM (e.g., Zephyr-7B or Mistral-7B)
-def load_llm():
+# Load a smaller LLM with customizable parameters
+def load_llm(temperature, top_k, max_length):
     model_name = "HuggingFaceH4/zephyr-7b-alpha"  # Replace with your preferred model
-    pipe = pipeline("text-generation", model=model_name, torch_dtype=torch.float16, device_map="auto")
+    pipe = pipeline(
+        "text-generation",
+        model=model_name,
+        torch_dtype=torch.float16,
+        device_map="auto",
+        temperature=temperature,
+        top_k=top_k,
+        max_length=max_length,
+    )
     llm = HuggingFacePipeline(pipeline=pipe)
     return llm
 
@@ -104,8 +112,14 @@ def main():
     # Create vector store
     vector_store = create_vector_store(chunks, indexing_method=indexing_method)
 
-    # Load LLM
-    llm = load_llm()
+    # LLM Parameters
+    st.sidebar.header("LLM Parameters")
+    temperature = st.sidebar.slider("Temperature", 0.1, 1.0, 0.7, help="Controls randomness in the output.")
+    top_k = st.sidebar.slider("Top-k", 1, 100, 50, help="Limits sampling to the top-k tokens.")
+    max_length = st.sidebar.slider("Max Length", 50, 500, 200, help="Maximum length of the generated response.")
+
+    # Load LLM with user-defined parameters
+    llm = load_llm(temperature=temperature, top_k=top_k, max_length=max_length)
 
     # Query translation options
     query_method = st.selectbox(