README.md

metadata

title: Ham Or Spam
emoji: 🔥
colorFrom: yellow
colorTo: gray
sdk: gradio
sdk_version: 4.14.0
app_file: app.py
pinned: false
license: apache-2.0

Ham or Spam Classifier with TensorFlow and Gradio

Project Overview

This project presents a machine learning solution for classifying text messages into 'Ham' (non-spam) or 'Spam'. It uses TensorFlow to build a Recurrent Neural Network (RNN) and deploys a Gradio interface for easy interaction with the model.

Key Technologies

• Python
• TensorFlow
• Gradio
• Pandas

Installation

To get started, ensure Python is installed on your system and then install the required libraries:

pip install pandas tensorflow gradio

Data Preprocessing

The dataset, assumed to be named 'spam.csv', is preprocessed as follows:

• Reading the CSV file.
• Converting the class labels to binary format (0 for 'ham', 1 for 'spam').
• Splitting the data into training and testing sets.

Model Architecture

The TensorFlow model includes:

• TextVectorization Layer: Processes the text data.
• Embedding Layer: Converts text to dense vector embeddings.
• Bidirectional LSTM Layer: Captures the context from both directions of text sequences.
• Dense Layers: For classification output.

Model Training

• The model is compiled with Binary Crossentropy loss and Adam optimizer.
• It is trained for 15 epochs on the training dataset and validated on the test dataset.

Model Evaluation

After training, the model is evaluated to determine its accuracy and loss:

test_loss, test_acc = model.evaluate(test_dataset)
print('Test Loss:', test_loss)
print('Test Accuracy:', test_acc)

Deployment with Gradio

The trained model is deployed using Gradio, which provides a user-friendly interface for real-time predictions:

iface = gr.Interface(
    fn=predict_spam, 
    inputs="text", 
    outputs="text",
    title="Ham or Spam Classifier",
    description="A Ham or Spam Classifier created using TensorFlow. Input a message to see if it's classified as Ham or Spam!"
)
iface.launch(share=True)

Running the Project

• Clone the project repository.
• Install the required dependencies.
• Execute the Python scripts for training the model and launching the Gradio interface.

Conclusion

This project demonstrates the use of TensorFlow and Gradio to build and deploy a practical machine learning solution for text classification. The model effectively distinguishes between 'Ham' and 'Spam' messages, making it a useful tool for email filtering or similar applications.