Create app.py

app.py

@@ -0,0 +1,57 @@
+import gradio as gr
+from langchain.vectorstores import Chroma
+from langchain.embeddings import HuggingFaceEmbeddings
+from langchain.llms import HuggingFaceHub
+from langchain.chains import ConversationalRetrievalChain
+from langchain.document_loaders import SimpleDocumentLoader
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain.memory import ConversationBufferMemory
+
+# Initialize the Hugging Face embedding model
+embedding_model = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
+
+# Initialize the LLaMA 2 model from Hugging Face Hub
+llm = HuggingFaceHub(repo_id="meta-llama/Llama-2-7b-hf", model_kwargs={"temperature": 0.7, "max_length": 512})
+
+# Initialize ChromaDB for storing and retrieving document embeddings
+vectorstore = Chroma(embedding_function=embedding_model, persist_directory="chroma_db")
+
+# Create a conversational chain with retrieval capabilities
+memory = ConversationBufferMemory(memory_key="chat_history", return_messages=True)
+qa_chain = ConversationalRetrievalChain.from_llm(llm, retriever=vectorstore.as_retriever(), memory=memory)
+
+def upload_docs(docs):
+    # Load and process the uploaded documents
+    loader = SimpleDocumentLoader(docs)
+    documents = loader.load()
+    
+    # Split documents into manageable chunks
+    text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
+    texts = text_splitter.split_documents(documents)
+    
+    # Add documents to the vector store and persist them
+    vectorstore.add_documents(texts)
+    vectorstore.persist()
+    
+    return "Documents uploaded and processed successfully!"
+
+def chat(query):
+    # Process the query with the conversational chain and return the result
+    response = qa_chain({"query": query})
+    return response["result"]
+
+# Gradio Interface
+with gr.Blocks() as demo:
+    with gr.Row():
+        with gr.Column():
+            doc_upload = gr.File(label="Upload your documents", file_types=[".txt", ".pdf", ".docx"], multiple=True)
+            upload_button = gr.Button("Upload")
+            upload_button.click(upload_docs, inputs=doc_upload, outputs=gr.Textbox())
+
+        with gr.Column():
+            chat_input = gr.Textbox(label="Ask a question:")
+            chat_output = gr.Textbox(label="Answer:")
+            chat_button = gr.Button("Send")
+            chat_button.click(chat, inputs=chat_input, outputs=chat_output)
+
+demo.launch()