app.py

#Import the necessary Libraries
import os
import uuid
import json
import gradio as gr
from openai import OpenAI
from langchain_community.embeddings.sentence_transformer import SentenceTransformerEmbeddings
from langchain_community.vectorstores import Chroma
from huggingface_hub import CommitScheduler
from pathlib import Path
from dotenv import load_dotenv

# Create Client
load_dotenv()

os.environ['OPENAI_API_KEY'] = anyscale_api_key

client = OpenAI(
    base_url="https://api.endpoints.anyscale.com/v1",
    api_key=os.environ['OPENAI_API_KEY']
)

# Define the embedding model and the vectorstore
embedding_model = SentenceTransformerEmbeddings(model_name='thenlper/gte-large')

# Load the persisted vectorDB
collection_name = 'picco-2023'

vectorstore_persisted = Chroma(
    collection_name=collection_name,
    embedding_function=embedding_model,
    persist_directory='/content/picco_db'
)

retriever = vectorstore_persisted.as_retriever(
    search_type="similarity",
    search_kwargs={'k': 5},
)

# Prepare the logging functionality
log_file = Path("logs/") / f"data_{uuid.uuid4()}.json"
log_folder = log_file.parent

scheduler = CommitScheduler(
    repo_id="picco-qna",
    repo_type="dataset",
    folder_path=log_folder,
    path_in_repo="data",
    every=2

# Define the Q&A system message
qna_system_message = """
You are an academic assistant who answers questions regarding pulse integrated continuous cardiac output (PiCCO) for learners preparing for PiCCO workshop .
User input will have the context required by you to answer user questions.
This context will begin with the token: ###Context.
The context contains references to specific portions of a document relevant to the user query.

User questions will begin with the token: ###Question.

Please answer only using the context provided in the input. Do not mention anything about the context in your final answer.

If the answer is not found in the context, respond "I don't know".
"""

# Define the user message template
qna_user_message_template = """
###Context
Here are some documents that are relevant to the question mentioned below.-
{context}

###Question
{question}
"""

# Define the predict function that runs when 'Submit' is clicked or when an API request is made
def predict(user_input):
    relevant_document_chunks = vectorstore_persisted.similarity_search(user_input, k=5)


    # Create context_for_query
    context_list = [f"Page {doc.metadata['page']}: {doc.page_content}" for doc in relevant_document_chunks]
    context_for_query = ".".join(context_list)

    # Create messages
    prompt = [
        {'role': 'system', 'content': qna_system_message},
        {'role': 'user', 'content': qna_user_message_template.format(context=context_for_query, question=user_input)}
    ]

    # Get response from the LLM
    try:
        response = client.chat.completions.create(
            model="mlabonne/NeuralHermes-2.5-Mistral-7B",
            messages=prompt,
            temperature=0
        )
        prediction = response.choices[0].message.content
    except Exception as e:
        prediction = f'Sorry, I encountered the following error: \n {e}'

    print(prediction)

    # Log both the inputs and outputs to a local log file
    with scheduler.lock:
        with log_file.open("a") as f:
            f.write(json.dumps({
                'user_input': user_input,
                'retrieved_context': context_for_query,
                'model_response': prediction
            }))
            f.write("\n")

    return prediction

# Set-up the Gradio UI

# Create the interface
demo = gr.Interface(
    inputs=gr.Textbox(placeholder="Enter your query here"),
    fn=predict,
    outputs="text",
    description="This web API presents an interface to ask questions on PiCCO Technology",
    article="Note that questions that are not relevant to PiCCO not be answered.",
    title="KSA/CCSK Hemodynamic Workshop Question & Answer Bot, by Chiko",
    concurrency_limit=16,
    examples = [
        ["State and explain the main scientific principles used in PiCCO technology"],
        ["What is the relevance of PiCCO in the operating room and critical care unit?"],
        ["Name and explain the parameters that can be measured using PiCCO, give their reference ranges"],
        ["State the basic equation that is used to estimate cardiac output using pulse contour analysis, explain each variable in detail"]

    ]
)

demo.queue()
demo.launch()