from langchain.document_loaders import HuggingFaceDatasetLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.vectorstores import FAISS
import gradio as gr


# Load the data
loader = HuggingFaceDatasetLoader("databricks/databricks-dolly-15k", "context")
data = loader.load()

# Document Transformers
text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=150)
docs = text_splitter.split_documents(data)

# Text Embedding
embeddings = HuggingFaceEmbeddings(
    model_name="sentence-transformers/all-MiniLM-l6-v2",
    model_kwargs={'device':'cpu'},
    encode_kwargs={'normalize_embeddings': False}
)

# Set up Vector Stores
db = FAISS.from_documents(docs, embeddings)

# Set up retrievers
retriever = db.as_retriever()


def generate(input):
    docs = retriever.get_relevant_documents(input)
    return docs[0].page_content


def respond(message, chat_history):
    bot_message = generate(message)
    chat_history.append((message, bot_message))
    
    return "", chat_history

# Set up the chat interface
with gr.Blocks() as demo:
    chatbot = gr.Chatbot(height=240) #just to fit the notebook
    msg = gr.Textbox(label="Ask away")
    btn = gr.Button("Submit")
    clear = gr.ClearButton(components=[msg, chatbot], value="Clear console")

    btn.click(respond, inputs=[msg, chatbot], outputs=[msg, chatbot])
    msg.submit(respond, inputs=[msg, chatbot], outputs=[msg, chatbot]) #Press enter to submit

demo.queue().launch()