app.py

import os
import streamlit as st
from together import Together
from langchain.vectorstores import Chroma
from langchain.embeddings import HuggingFaceBgeEmbeddings
from langchain.chains import ConversationalRetrievalChain

# --- Configuration ---
TOGETHER_API_KEY = os.environ.get("TOGETHER_API_KEY")
if not TOGETHER_API_KEY:
    st.error("Missing TOGETHER_API_KEY environment variable.")
    st.stop()

# Initialize TogetherAI client
client = Together(api_key=TOGETHER_API_KEY)

# Embeddings setup
EMBED_MODEL_NAME = "BAAI/bge-base-en"
embeddings = HuggingFaceBgeEmbeddings(
    model_name=EMBED_MODEL_NAME,
    encode_kwargs={"normalize_embeddings": True},
)

# Sidebar: select collection
st.sidebar.title("DocChatter RAG")
collection = st.sidebar.selectbox(
    "Choose a document collection:",
    ['General Medicine', 'RespiratoryFishman', 'RespiratoryMurray', 'MedMRCP2', 'OldMedicine']
)

dirs = {
    'General Medicine': './oxfordmedbookdir/',
    'RespiratoryFishman': './respfishmandbcud/',
    'RespiratoryMurray': './respmurray/',
    'MedMRCP2': './medmrcp2store/',
    'OldMedicine': './mrcpchromadb/'
}
cols = {
    'General Medicine': 'oxfordmed',
    'RespiratoryFishman': 'fishmannotescud',
    'RespiratoryMurray': 'respmurraynotes',
    'MedMRCP2': 'medmrcp2notes',
    'OldMedicine': 'mrcppassmednotes'
}

persist_directory = dirs.get(collection)
collection_name = cols.get(collection)

# Load Chroma vector store
vectorstore = Chroma(
    collection_name=collection_name,
    persist_directory=persist_directory,
    embedding_function=embeddings
)
retriever = vectorstore.as_retriever(search_kwargs={"k":5})

# System prompt template
SYSTEM_PROMPT = (
    "You are a helpful assistant for medical professionals. "
    "Use the following context from medical documents to answer the question. "
    "If you don't know, say you don't know.\n\nContext:\n{context}\n"
)

st.title("🩺 DocChatter RAG (Streaming)")

# Initialize chat history
if 'chat_history' not in st.session_state:
    st.session_state.chat_history = []  # list of dicts {role, content}

# Tabs
chat_tab, clear_tab = st.tabs(["Chat", "Clear History"])
with chat_tab:
    # Display history
    for msg in st.session_state.chat_history:
        st.chat_message(msg['role']).write(msg['content'])

    # User input
    if prompt := st.chat_input("Ask anything about your docs..."):
        # User message
        st.chat_message("user").write(prompt)
        st.session_state.chat_history.append({"role": "user", "content": prompt})

        # Retrieve relevant docs
        docs = retriever.get_relevant_documents(prompt)
        context = "\n---\n".join([d.page_content for d in docs])

        # Build messages for TogetherAI
        system_msg = {"role": "system", "content": SYSTEM_PROMPT.format(context=context)}
        messages = [system_msg]
        # include prior conversation
        for msg in st.session_state.chat_history:
            if msg['role'] in ('user', 'assistant'):
                messages.append(msg)
        # Prepare streaming response
        response_container = st.chat_message("assistant")
        placeholder = response_container.empty()
        answer = ""
        # Stream tokens
        for token in client.chat.completions.create(
            model="meta-llama/Llama-4-Scout-17B-16E-Instruct",
            messages=messages,
            stream=True
        ):
            if hasattr(token, 'choices'):
                delta = token.choices[0].delta.content
                answer += delta
                placeholder.write(answer)
        # Save assistant message
        st.session_state.chat_history.append({"role": "assistant", "content": answer})

with clear_tab:
    if st.button("🗑️ Clear chat history"):
        st.session_state.chat_history = []
        st.experimental_rerun()