app.py

import gradio as gr
from transformers import pipeline

# Load AI models
caption_generator = pipeline("text-generation", model="Xenova/distilgpt2")
summarizer = pipeline("summarization", model="Xenova/t5-small")
sentiment_analyzer = pipeline("sentiment-analysis", model="Xenova/distilbert-base-uncased-finetuned-sst-2-english")

# Functions
def generate_caption(text):
    result = caption_generator(text, max_length=50)
    return result[0]['generated_text']

def summarize_text(text):
    result = summarizer(text, max_length=60, min_length=20, do_sample=False)
    return result[0]['summary_text']

def analyze_sentiment(text):
    result = sentiment_analyzer(text)
    return f"{result[0]['label']} ({result[0]['score']:.2f})"

# Gradio UI
with gr.Blocks() as demo:
    gr.Markdown("# AI Tools Website\nGenerate captions, summarize text, analyze sentiment!")
    with gr.Tab("Generate Caption"):
        inp1 = gr.Textbox(label="Enter text")
        out1 = gr.Textbox(label="Caption")
        btn1 = gr.Button("Generate Caption")
        btn1.click(generate_caption, inp1, out1)
    with gr.Tab("Summarize"):
        inp2 = gr.Textbox(label="Enter text")
        out2 = gr.Textbox(label="Summary")
        btn2 = gr.Button("Summarize")
        btn2.click(summarize_text, inp2, out2)
    with gr.Tab("Sentiment"):
        inp3 = gr.Textbox(label="Enter text")
        out3 = gr.Textbox(label="Sentiment")
        btn3 = gr.Button("Analyze Sentiment")
        btn3.click(analyze_sentiment, inp3, out3)

demo.launch()