app.py

import gradio as gr
from sentence_transformers import SentenceTransformer, util
# import openai
from groq import Groq #FOR GROQ API KEY
import os

os.environ["TOKENIZERS_PARALLELISM"] = "false"

# Initialize paths and model identifiers for easy configuration and maintenance
filename = "output_topic_details.txt"  # Path to the file storing chess-specific details
retrieval_model_name = 'output/sentence-transformer-finetuned/'

# openai.api_key = os.environ["OPENAI_API_KEY"]
client = Groq(api_key=os.environ.get("GROQ_API_KEY")) #FOR GROQ API KEY

system_message = "You are a South Korean food chatbot specialized in providing information on korean foods' cultural significance, ingredients, eating etiquette, and/or similar tasting foods. You are of Korean origin and know about the food and culture very well. You should sound like a tour guide and be super excited to share your expertise in Korean food (you have a PhD in Korean gastronomy). But you are also fluent in English and have lived in the US. However, your answer should always be concise on only answer the question being answered. Always give responses under 500 tokens"
# Initial system message to set the behavior of the assistant
messages = [{"role": "system", "content": system_message}]

# Attempt to load the necessary models and provide feedback on success or failure
try:
    retrieval_model = SentenceTransformer(retrieval_model_name)
    print("Models loaded successfully.")
except Exception as e:
    print(f"Failed to load models: {e}")

def load_and_preprocess_text(filename):
    """
    Load and preprocess text from a file, removing empty lines and stripping whitespace.
    """
    try:
        with open(filename, 'r', encoding='utf-8') as file:
            segments = [line.strip() for line in file if line.strip()]
        print("Text loaded and preprocessed successfully.")
        return segments
    except Exception as e:
        print(f"Failed to load or preprocess text: {e}")
        return []

segments = load_and_preprocess_text(filename)

def find_relevant_segment(user_query, segments):
    """
    Find the most relevant text segment for a user's query using cosine similarity among sentence embeddings.
    This version finds the best match based on the content of the query.
    """
    try:
        # Lowercase the query for better matching
        lower_query = user_query.lower()
        
        # Encode the query and the segments
        query_embedding = retrieval_model.encode(lower_query)
        segment_embeddings = retrieval_model.encode(segments)
        
        # Compute cosine similarities between the query and the segments
        similarities = util.pytorch_cos_sim(query_embedding, segment_embeddings)[0]
        
        # Find the index of the most similar segment
        best_idx = similarities.argmax()
        
        # Return the most relevant segment
        return segments[best_idx]
    except Exception as e:
        print(f"Error in finding relevant segment: {e}")
        return ""

def generate_response(user_query, relevant_segment):
    """
    Generate a response emphasizing the bot's capability in providing information about Korean food.
    """
    try:
        user_message = f"Here's the information on Korean Food: {relevant_segment}"

        # Append user's message to messages list
        messages.append({"role": "user", "content": user_message})
        
        # response = openai.ChatCompletion.create(
        #     model="gpt-4o",
        #     messages=messages,
        #     max_tokens=700,
        #     temperature=0.5,
        #     top_p=1,
        #     frequency_penalty=0.5,
        #     presence_penalty=0.5
        # )
        response = client.chat.completions.create( #FOR GROK API KEY
            model="llama-3.3-70b-versatile", #changed model
            messages=messages,
            max_tokens=500,
            temperature=0.5
        )

        
        # Extract the response text
        # output_text = response['choices'][0]['message']['content'].strip()
        output_text = response.choices[0].message.content.strip() #FOR GROQ API KEY
        
        # Append assistant's message to messages list for context
        messages.append({"role": "assistant", "content": output_text})
        
        return output_text
        
    except Exception as e:
        print(f"Error in generating response: {e}")
        return f"Error in generating response: {e}"

def query_model(question):
    """
    Process a question, find relevant information, and generate a response.
    """
    if question == "":
        return "Welcome to Globowl! Ask me anything about Korean foods' cultural significance, ingredients, eating etiquette, and/or similar tasting foods."
    relevant_segment = find_relevant_segment(question, segments)
    if not relevant_segment:
        return "Could not find specific information. Please refine your question."
    response = generate_response(question, relevant_segment)
    return response

# Define the welcome message and specific topics the chatbot can provide information about


topics = """

<span>

"Choose one of the food names and what you want to learn about the food from the "fact options" section below!
\nfood options
\n- Kimbap
\n- Kimchi
\n- Tofu Soup
\n- Bibimbap
\n- Seaweed Snack

\nfact options
\n- ingredients
\n- history behind it
\n- eating etiquette
\n- similar food around the world 

</span>
"""


# Setup the Gradio Blocks interface with custom layout components
with gr.Blocks(theme='JohnSmith9982/small_and_pretty') as demo:
  # Display the formatted welcome message
    gr.Image("Globowl.png", show_label = False, show_share_button = False, show_download_button = False)
    with gr.Row():
        with gr.Column():
            gr.Markdown(topics)  # Show the topics on the left side
    # with gr.Row():
    #    with gr.Column():
    #        question = gr.Textbox(label="Your question", placeholder="What do you want to ask about?")
    #        answer = gr.Textbox(label="GloBowl Response [Korean]", placeholder="GloBowl will respond here...", interactive=False, lines=10, show_copy_button = True)
    #        submit_button = gr.Button("Submit")
    #        submit_button.click(fn=query_model, inputs=question, outputs=answer)
    
    chatbot = gr.Interface(
        
            fn=query_model,
            inputs=gr.Textbox(label="Your question", placeholder="What do you want to ask about?"),
            outputs=gr.Textbox(label="GloBowl Response [Korean]", placeholder="GloBowl will respond here...", interactive=False, lines=10, show_copy_button = True
),
          #   fn=query_model,
          #   #title=title,
          # #  description=description,
          #  # examples=examples,
          #   inputs=["Your question", "state"],
          #   outputs=["chatbot", "state"],
    )
    


# Launch the Gradio app to allow user interaction
demo.launch(share=True)