app.py

import gradio as gr
from huggingface_hub import InferenceClient
from transformers import pipeline
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer

# Load the chatbot model and tokenizer
pvm_model_adapt = "facebook/blenderbot-400M-distill"
tokenizer = AutoTokenizer.from_pretrained(pvm_model_adapt)
model = AutoModelForSeq2SeqLM.from_pretrained(pvm_model_adapt)


from textblob import TextBlob

def analyze_sentiment(text):
    sentiment_score = TextBlob(text).sentiment.polarity  # Score: -1 (negative) to +1 (positive)
    if sentiment_score > 0.3:
        return "positive"
    elif sentiment_score < -0.3:
        return "negative"
    else:
        return "neutral"

def chatbot_response(user_input):
    sentiment = analyze_sentiment(user_input)
    
    # Generate a chatbot response
    inputs = tokenizer(user_input, return_tensors="pt")
    output = model.generate(**inputs)
    response = tokenizer.decode(output[0], skip_special_tokens=True)

    # Adaptive UI modifications
    if sentiment == "positive":
        ui_style = "😊 Friendly mode activated!"
    elif sentiment == "negative":
        ui_style = "💙 Supportive mode activated. I'm here to help."
    else:
        ui_style = "🙂 Neutral mode."

    return f"{ui_style}\n\nChatbot: {response}"


    import gradio as gr

# Set up the Gradio UI
iface = gr.Interface(
    fn=chatbot_response,
    inputs=gr.Textbox(lines=2, placeholder="start with your message..."),
    outputs="text",
    title="PVM Adaptive Chat",
    description="This chatbot adapts to the user sentiment.",
)

iface.launch()