app.py

import gradio as gr
from transformers import AutoModelForSequenceClassification, AutoTokenizer
import torch
import torch.nn.functional as F
import logging
from datetime import datetime

# Load model and tokenizer
model_path = "./" 
model = AutoModelForSequenceClassification.from_pretrained(model_path)
tokenizer = AutoTokenizer.from_pretrained(model_path)

# Configure logging
LOG_FILE = "user_logs.txt"
logging.basicConfig(
    filename=LOG_FILE,
    level=logging.INFO,
    format="%(asctime)s - %(levelname)s - %(message)s",
)

def log_user_input(input_text: str, source: str = "WebApp"):
    log_entry = f"{datetime.now()} | Source: {source} | Input: {input_text}"
    logging.info(log_entry)

def predict_sentiment(text: str) -> dict:
    log_user_input(text)
    inputs = tokenizer(
        text,
        return_tensors="pt",
        padding="max_length",
        truncation=True,
        max_length=128,
    )
    with torch.no_grad():
        outputs = model(**inputs)
    probs = F.softmax(outputs.logits, dim=-1)
    sentiment_probs = {
        "Neutral": round(probs[0][0].item(), 4),
        "Positive": round(probs[0][1].item(), 4),
        "Negative": round(probs[0][2].item(), 4),
    }
    return sentiment_probs

# Gradio Interface
iface = gr.Interface(
    fn=predict_sentiment,
    inputs=gr.Textbox(
        lines=2,
        placeholder="Type a tweet or sentence here...",
        label="Enter Text",
        elem_id="input_textbox"
    ),
    outputs=gr.JSON(label="Sentiment Probabilities"),
    title="Sentiment Analysis Application",
    description="Analyze text sentiment (Negative, Neutral, Positive) with probabilities.",
    examples=[
        "I love this product!",
        "This is the worst service I've ever had.",
        "The weather today is neutral, not too hot or cold.",
    ],
    css="""  /* Styling for Input Textbox */
    #input_textbox {
        font-size: 16px;
        padding: 15px;
        border-radius: 10px;
        border: 1px solid #4CAF50;  /* Green border */
        width: 80%;
        margin: 0 auto;
        font-family: 'Arial', sans-serif;
        background-color: #f4fdf1;  /* Light green background */
    }
    /* Styling for the entire Gradio container */
    .gradio-container {
        background-color: #f0f8ff;  /* Light blue background */
        border-radius: 15px;
        box-shadow: 0 4px 10px rgba(0, 0, 0, 0.1);
        padding: 20px;
    }
    /* Title Styling */
    .gradio-title {
        color: #2C3E50;  /* Dark gray-blue text color */
        font-family: 'Arial', sans-serif;
        font-size: 28px;
        font-weight: bold;
        text-align: center;
    }
    /* Description Styling */
    .gradio-description {
        font-size: 16px;
        font-family: 'Arial', sans-serif;
        color: #34495E;  /* Darker gray color */
        text-align: center;
    }
    /* Button Styling */
    .gradio-btn {
        background-color: #4CAF50;  /* Green button */
        color: white;
        border-radius: 8px;
        padding: 10px;
        font-size: 14px;
        width: 100%;
        font-family: 'Arial', sans-serif;
        margin-top: 15px;
    }
    /* Button Hover Styling */
    .gradio-btn:hover {
        background-color: #45a049;  /* Slightly darker green on hover */
    }
    /* Output JSON Styling */
    .gradio-json {
        font-size: 16px;
        color: #333;
        background-color: #eaf2f8;  /* Very light blue background */
        padding: 10px;
        border-radius: 8px;
        border: 1px solid #ddd;
    }
    """
)

iface.launch(share=True)