Create app.py

app.py

@@ -0,0 +1,111 @@
+import os
+import sys
+import gradio.inputs as gr_inputs
+import gradio.outputs as gr_outputs
+import gradio as gr
+from pinecone import Pinecone, ServerlessSpec
+from langchain_community.llms import Replicate
+from langchain_pinecone import PineconeVectorStore
+from langchain.text_splitter import CharacterTextSplitter
+from langchain_community.document_loaders import PyPDFLoader
+from langchain_huggingface.embeddings import HuggingFaceEmbeddings
+from langchain.chains import ConversationalRetrievalChain
+import time
+
+key1 = os.environ.get('REPLICATE_API_TOKEN')
+key2 = os.environ.get('PINECONE_API_KEY')
+os.environ['REPLICATE_API_TOKEN'] = key1
+os.environ["PINECONE_API_KEY"] = key2
+
+# Initialize Pinecone
+pc = Pinecone(api_key=os.environ["PINECONE_API_KEY"])
+
+# Function to process PDF and set up chatbot
+def process_pdf(pdf_doc):
+    # Save uploaded file
+    filename = pdf_doc.name
+    pdf_doc.save(filename)
+
+    # Load PDF and create index
+    loader = PyPDFLoader(filename)
+    documents = loader.load()
+    text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
+    texts = text_splitter.split_documents(documents)
+    
+    embeddings = HuggingFaceEmbeddings()
+
+    index_name = "pdfchatbot"
+    existing_indexes = [index_info["name"] for index_info in pc.list_indexes()]
+
+    if index_name in existing_indexes:
+        pc.delete_index(index_name)
+        while index_name in [index_info["name"] for index_info in pc.list_indexes()]:
+            time.sleep(1)
+
+    pc.create_index(
+        name=index_name,
+        dimension=768,
+        metric="cosine",
+        spec=ServerlessSpec(cloud="aws", region="us-east-1"),
+    )
+    while not pc.describe_index(index_name).status["ready"]:
+        time.sleep(1)
+
+    index = pc.Index(index_name)
+
+    vectordb = PineconeVectorStore.from_documents(texts, embeddings, index_name=index_name)
+
+    llm = Replicate(
+        model="a16z-infra/llama13b-v2-chat:df7690f1994d94e96ad9d568eac121aecf50684a0b0963b25a41cc40061269e5",
+        input={"temperature": 0.75, "max_length": 3000}
+    )
+
+    global qa_chain
+    qa_chain = ConversationalRetrievalChain.from_llm(
+        llm,
+        vectordb.as_retriever(search_kwargs={'k': 2}),
+        return_source_documents=True
+    )
+
+    return "Ready"
+
+# Function to handle user queries
+def query(history, text):
+    langchain_history = [(msg[1], history[i+1][1] if i+1 < len(history) else "") for i, msg in enumerate(history) if i % 2 == 0]
+    result = qa_chain({"question": text, "chat_history": langchain_history})
+    new_history = history + [(text,result['answer'])]
+    return new_history, ""
+
+# Define the Gradio interface
+css = """
+#col-container {max-width: 700px; margin-left: auto; margin-right: auto;}
+"""
+
+title_html = """
+<div style="text-align: center;max-width: 700px;">
+    <h1>Chat with PDF</h1>   
+"""
+
+iface = gr.Interface(
+    fn=process_pdf,
+    inputs=gr_inputs.File(label="Load a PDF", type="file", accept=".pdf"),
+    outputs=gr_outputs.Textbox(label="Status", type="auto", default=""),
+    title="PDF Chatbot Interface",
+    description="Upload a PDF file to start interacting with the chatbot.",
+    allow_flagging=False,
+    css=css
+)
+
+# Add chat history and question input to the interface
+chatbot_interface = gr.Interface(
+    fn=query,
+    inputs=gr_inputs.Textbox(label="Question", placeholder="Type your question and hit Enter"),
+    outputs=gr_outputs.Textbox(label="Chat History", type="auto", default=""),
+    title=title_html,
+    live=True,
+    css=css
+)
+
+# Launch the combined interface
+iface.launch()
+chatbot_interface.launch()