remove dotenv

app.py

@@ -2,7 +2,6 @@ import os
 import io
 import requests
 import streamlit as st
-from dotenv import load_dotenv
 from PyPDF2 import PdfReader
 from langchain.text_splitter import CharacterTextSplitter
 from langchain.embeddings.huggingface import HuggingFaceEmbeddings
@@ -11,12 +10,6 @@ from langchain.chains.question_answering import load_qa_chain
 from langchain.llms import HuggingFacePipeline
 from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
 
-# Disable WANDB
-os.environ['WANDB_DISABLED'] = "true"
-
-# Constants
-MODEL_PATH = "/home/lab/halyn/gemma/halyn/paper/models/gemma-2-9b-it"
-
 # Global variables
 knowledge_base = None
 qa_chain = None
@@ -24,10 +17,6 @@ qa_chain = None
 def load_pdf(pdf_file):
     """
     Load and extract text from a PDF.
-    Args:
-        pdf_file (str): The PDF file.
-    Returns:
-        str: Extracted text from the PDF.
     """
     pdf_reader = PdfReader(pdf_file)
     text = "".join(page.extract_text() for page in pdf_reader.pages)
@@ -36,10 +25,6 @@ def load_pdf(pdf_file):
 def split_text(text):
     """
     Split the extracted text into chunks.
-    Args:
-        text (str): The full text extracted from the PDF.
-    Returns:
-        list: A list of text chunks.
     """
     text_splitter = CharacterTextSplitter(
         separator="\n", chunk_size=1000, chunk_overlap=200, length_function=len
@@ -49,10 +34,6 @@ def split_text(text):
 def create_knowledge_base(chunks):
     """
     Create a FAISS knowledge base from text chunks.
-    Args:
-        chunks (list): A list of text chunks.
-    Returns:
-        FAISS: A FAISS knowledge base object.
     """
     embeddings = HuggingFaceEmbeddings()
     return FAISS.from_texts(chunks, embeddings)
@@ -60,10 +41,6 @@ def create_knowledge_base(chunks):
 def load_model(model_path):
     """
     Load the HuggingFace model and tokenizer, and create a text-generation pipeline.
-    Args:
-        model_path (str): The path to the pre-trained model.
-    Returns:
-        pipeline: A HuggingFace pipeline for text generation.
     """
     tokenizer = AutoTokenizer.from_pretrained(model_path)
     model = AutoModelForCausalLM.from_pretrained(model_path)

requirements.txt

@@ -5,4 +5,4 @@ transformers==4.31.0
 torch==2.0.1
 faiss-cpu==1.7.4
 requests==2.31.0
-huggingface-hub==0.16.4
+huggingface-hub==0.16.4