Update app.py

app.py

@@ -19,14 +19,12 @@ def process_answer(instruction, qa_chain):
     generated_text = qa_chain.run(instruction)
     return generated_text
 
-
 def get_file_size(file):
     file.seek(0, os.SEEK_END)
     file_size = file.tell()
     file.seek(0)
     return file_size
 
-
 @st.cache_resource
 def data_ingestion():
     for root, dirs, files in os.walk("docs"):
@@ -39,15 +37,14 @@ def data_ingestion():
     text_splitter = RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=500)
     splits = text_splitter.split_documents(documents)
 
-    # create embeddings here
+    # Hier Embeddings erstellen
     embeddings = SentenceTransformerEmbeddings(model_name="all-MiniLM-L6-v2")
     vectordb = FAISS.from_documents(splits, embeddings)
     vectordb.save_local("faiss_index")
 
-
 @st.cache_resource
 def initialize_qa_chain(selected_model):
-    # Constants
+    # Konstanten
     CHECKPOINT = selected_model
     TOKENIZER = AutoTokenizer.from_pretrained(CHECKPOINT)
     BASE_MODEL = AutoModelForSeq2SeqLM.from_pretrained(CHECKPOINT, device_map=torch.device('cpu'), torch_dtype=torch.float32)
@@ -67,7 +64,7 @@ def initialize_qa_chain(selected_model):
 
     vectordb = FAISS.load_local("faiss_index", embeddings)
 
-    # Build a QA chain
+    # QA-Kette erstellen
     qa_chain = RetrievalQA.from_chain_type(
         llm=llm,
         chain_type="stuff",
@@ -75,45 +72,42 @@ def initialize_qa_chain(selected_model):
     )
     return qa_chain
 
-
 @st.cache_data
-# function to display the PDF of a given file
+# Funktion zum Anzeigen der PDF einer bestimmten Datei
 def display_pdf(file):
     try:
-        # Opening file from file path
+        # Datei von Dateipfad öffnen
         with open(file, "rb") as f:
             base64_pdf = base64.b64encode(f.read()).decode('utf-8')
 
-        # Embedding PDF in HTML
+        # PDF in HTML einbetten
         pdf_display = f'<iframe src="data:application/pdf;base64,{base64_pdf}" width="100%" height="600" type="application/pdf"></iframe>'
 
-        # Displaying File
+        # Datei anzeigen
         st.markdown(pdf_display, unsafe_allow_html=True)
     except Exception as e:
-        st.error(f"An error occurred while displaying the PDF: {e}")
+        st.error(f"Ein Fehler ist beim Anzeigen der PDF aufgetreten: {e}")
 
-
-# Display conversation history using Streamlit messages
+# Unterhaltungsgeschichte mit Streamlit-Nachrichten anzeigen
 def display_conversation(history):
     for i in range(len(history["generated"])):
         message(history["past"][i], is_user=True, key=f"{i}_user")
         message(history["generated"][i], key=str(i))
 
-
 def main():
-     # Add a sidebar for model selection
+    # Sidebar für die Modellauswahl hinzufügen
     model_options = ["MBZUAI/LaMini-T5-738M", "google/flan-t5-base", "google/flan-t5-small"]
-    selected_model = st.sidebar.selectbox("Select Model", model_options)
+    selected_model = st.sidebar.selectbox("Modell auswählen", model_options)
     
-    st.markdown("<h1 style='text-align: center; color: blue;'>Custom PDF Chatbot 🦜📄 </h1>", unsafe_allow_html=True)
-    st.markdown("<h2 style='text-align: center; color:red;'>Upload your PDF, and Ask Questions 👇</h2>", unsafe_allow_html=True)
+    st.markdown("<h1 style='text-align: center; color: blue;'>Benutzerdefinierter PDF-Chatbot 🦜📄 </h1>", unsafe_allow_html=True)
+    st.markdown("<h2 style='text-align: center; color:red;'>Laden Sie Ihr PDF hoch und stellen Sie Fragen 👇</h2>", unsafe_allow_html=True)
 
     uploaded_file = st.file_uploader("", type=["pdf"])
 
     if uploaded_file is not None:
         file_details = {
-            "Filename": uploaded_file.name,
-            "File size": get_file_size(uploaded_file)
+            "Dateiname": uploaded_file.name,
+            "Dateigröße": get_file_size(uploaded_file)
         }
         os.makedirs("docs", exist_ok=True)
         filepath = os.path.join("docs", uploaded_file.name)
@@ -123,40 +117,39 @@ def main():
 
             col1, col2 = st.columns([1, 2])
             with col1:
-                st.markdown("<h4 style color:black;'>File details</h4>", unsafe_allow_html=True)
+                st.markdown("<h4 style color:black;'>Dateidetails</h4>", unsafe_allow_html=True)
                 st.json(file_details)
-                st.markdown("<h4 style color:black;'>File preview</h4>", unsafe_allow_html=True)
+                st.markdown("<h4 style color:black;'>Dateivorschau</h4>", unsafe_allow_html=True)
                 pdf_view = display_pdf(filepath)
 
             with col2:
-                st.success(f'model selected successfully: {selected_model}')
-                with st.spinner('Embeddings are in process...'):
+                st.success(f'Modell erfolgreich ausgewählt: {selected_model}')
+                with st.spinner('Embeddings werden erstellt...'):
                     ingested_data = data_ingestion()
-                st.success('Embeddings are created successfully!')
-                st.markdown("<h4 style color:black;'>Chat Here</h4>", unsafe_allow_html=True)
+                st.success('Embeddings wurden erfolgreich erstellt!')
+                st.markdown("<h4 style color:black;'>Hier chatten</h4>", unsafe_allow_html=True)
 
                 user_input = st.text_input("", key="input")
 
-                # Initialize session state for generated responses and past messages
+                # Sitzungszustand für generierte Antworten und vergangene Nachrichten initialisieren
                 if "generated" not in st.session_state:
-                    st.session_state["generated"] = ["I am ready to help you"]
+                    st.session_state["generated"] = ["Ich bin bereit, Ihnen zu helfen"]
                 if "past" not in st.session_state:
-                    st.session_state["past"] = ["Hey there!"]
+                    st.session_state["past"] = ["Hallo!"]
 
-                # Search the database for a response based on user input and update session state
+                # In der Datenbank nach einer Antwort basierend auf der Benutzereingabe suchen und den Sitzungszustand aktualisieren
                 if user_input:
                     answer = process_answer({'query': user_input}, initialize_qa_chain(selected_model))
                     st.session_state["past"].append(user_input)
                     response = answer
                     st.session_state["generated"].append(response)
 
-                # Display conversation history using Streamlit messages
+                # Unterhaltungsgeschichte mit Streamlit-Nachrichten anzeigen
                 if st.session_state["generated"]:
                     display_conversation(st.session_state)
 
         except Exception as e:
-            st.error(f"An error occurred: {e}")
-
+            st.error(f"Ein Fehler ist aufgetreten: {e}")
 
 if __name__ == "__main__":
     main()