Update app.py

app.py

@@ -13,6 +13,7 @@ from langchain.prompts import PromptTemplate
 from langchain_community.document_loaders import PyPDFLoader
 from langchain_chroma import Chroma
 from langchain_community.vectorstores import Chroma
+from transformers import AutoTokenizer, AutoModelForCausalLM
 
 #st.set_page_config(page_title="Document Genie", layout="wide")
 
@@ -60,7 +61,7 @@ def text_splitter(text):
 
 #GOOGLE_API_KEY = os.getenv("GOOGLE_API_KEY")
 #COHERE_API_KEY = os.getenv("COHERE_API_KEY")
-HUGGING_FACE_API_KEY = os.getenv("HUGGINGFACE_API_KEY")
+#HUGGING_FACE_API_KEY = os.getenv("HUGGINGFACE_API_KEY")
 
 def get_conversational_chain():
     prompt_template = """
@@ -78,8 +79,11 @@ def get_conversational_chain():
     #model = ChatGoogleGenerativeAI(model="gemini-pro", temperature=0.3, google_api_key=GOOGLE_API_KEY)
     repo_id ='google/gemma-1.1-2b-it'
     #repo_id='meta-llama/Meta-Llama-3-70B'
-    llm = HuggingFaceEndpoint(
-    repo_id=repo_id, max_length=512, temperature=0.5, token=HUGGING_FACE_API_KEY)
+    #llm = HuggingFaceEndpoint(
+    #repo_id=repo_id, max_length=512, temperature=0.5, token=HUGGING_FACE_API_KEY)
+    tokenizer = AutoTokenizer.from_pretrained("google/gemma-1.1-2b-it")
+    llm = AutoModelForCausalLM.from_pretrained("google/gemma-1.1-2b-it")
+
     pt = ChatPromptTemplate.from_template(template)
     # Retrieve and generate using the relevant snippets of the blog.
     retriever = db.as_retriever()