🧠 Intent Classifier using Transformers

A multi-class text classification project fine-tuned on a custom dataset of user queries from a simulated customer support domain. This project demonstrates end-to-end fine-tuning of a transformer model (DistilBERT) for intent detection.

---

📌 Project Overview

This project performs intent classification on a labeled dataset of customer queries. It involves:

• Data cleaning and preprocessing
• Baseline modeling with TF-IDF + Logistic Regression
• Fine-tuning a pre-trained transformer (DistilBERT)
• Model evaluation using accuracy, precision, recall, and F1-score
• Visualization of confusion matrix

---

🧾 Dataset

• Total Samples: 600 (synthetically generated and paraphrased)
• Labels (5 classes):
  • product_quality
  • delivery_issue
  • price_concern
  • return_request
  • general_inquiry

The dataset is stored in the data/ folder with separate train/test splits.

---

🧰 Tech Stack

• Python 3.10+
• PyTorch
• Hugging Face Transformers
• Scikit-learn
• Pandas, NumPy
• Matplotlib, Seaborn

---

🚀 How to Run

1. Clone this repo:

   git clone https://github.com/rehan-shafi/intent-classifier.git
   cd intent-classifier
   

2. Install dependencies: pip install -r requirements.txt

3. Train the model: python train_model.py

4. Outputs: Trained model saved in ./model_output Classification report and confusion matrix printed after training

📊 Sample Evaluation Output: Accuracy: 0.72 Macro F1 Score: 0.73

Confusion Matrix: (See plot in output section of notebook/script)

📁 Project Structure intent-classifier/ ├── data/ │ ├── train.csv │ └── test.csv ├── notebooks/ │ ├── tfidf_baseline.ipynb │ └── transformer_finetune.ipynb ├── scripts/ │ ├── model_utils.py │ └── train_model.py ├── model_output/ ├── README.md ├── requirements.txt └── LICENSE

📦 Future Improvements Hyperparameter tuning with Optuna Add model card and Hugging Face deployment Export model to ONNX or TorchScript for inference Streamlit or Gradio UI for demo

🧑‍💻 Author Rehan Shafi Generative AI Developer GitHub Profile (update with your GitHub)

📜 License MIT License – feel free to fork, modify, and use for educational purposes.