Contextual Chatbot Implementation in PyTorch

Project Description

This project presents an advanced implementation of a contextual chatbot using PyTorch. The chatbot leverages deep learning techniques to understand and respond to user inputs in a conversational manner. The project is designed to be beginner-friendly while providing a comprehensive understanding of chatbot development. It includes a straightforward Feed Forward Neural Network with two hidden layers and allows for easy customization to fit various use cases.

Sample Outputs

Question: Can you help me with something?
Response: Sure. Tell me how can i assist you
--------------------------------------------------
Question: Can you help me with something?
Response: Tell me your problem so that i can assist you
--------------------------------------------------
Question: Can you help me with something?
Response: Yes, sure. How can I help you?
--------------------------------------------------
Question: I am feeling lonely
Response: Why do you think you feel this way?
--------------------------------------------------
Question: I feel so worthless.
Response: I'm trying my best to help you. So please talk to me
--------------------------------------------------
Question: I feel so worthless.
Response: I'm sorry to hear that. I'm doing my best to help
--------------------------------------------------
Question: I feel so anxius.
Response: Did something happen which made you feel this way?
--------------------------------------------------
Question: I can't take it anymore
Response: I'm trying my best to help you. So please talk to me
--------------------------------------------------
Question: I am so stressed out
Response: Give yourself a break. Go easy on yourself.
--------------------------------------------------
Question: I am so stressed out
Response: What do you think is causing this?
--------------------------------------------------
Question: I am so stressed out
Response: Give yourself a break. Go easy on yourself.
--------------------------------------------------
Question: I have insominia
Response: That seem awful. What do you think is behind this?
--------------------------------------------------
Question: I'm scared
Response: That's geat to hear. I'm glad you're feeling this way.
--------------------------------------------------
Question: I'm scared
Response: Oh i see. That's great.
--------------------------------------------------
Question: I'm scared
Response: That's geat to hear. I'm glad you're feeling this way.
--------------------------------------------------
Question: My mom died
Response: I'm sorry to hear that. If you want to talk about it. I'm here.
--------------------------------------------------
Question: My mom died
Response: I am really sorry to hear that. I am here to help you with grief, anxiety and anything else you may feel at this time.
--------------------------------------------------
Question: My mom died
Response: My condolences. I'm here if you need to talk.
--------------------------------------------------
Question: You don't understand me.
Response: I'm sorry to hear that. I'm doing my best to help
--------------------------------------------------
Question: You don't understand me.
Response: I'm sorry to hear that. I'm doing my best to help
--------------------------------------------------
Question: You don't understand me.
Response: It sound like i'm not being very helpful right now.
--------------------------------------------------
Question: That's all.
Response: I'm listening. Please go on.
--------------------------------------------------
Question: That's all.
Response: Come elucidate your thoughts
--------------------------------------------------
Question: That's all.
Response: Let's discuss further why you're feeling this way.
--------------------------------------------------

Technologies and Techniques Used

• Programming Language: Python
• Deep Learning Framework: PyTorch
• Natural Language Processing: NLTK (Natural Language Toolkit)
• Model Architecture: Feed Forward Neural Network
• Training Data: Customizable intents defined in data.json
• Tokenization and Vocabulary: Custom tokenizer and vocabulary files

How to Run the Project

Prerequisites

Ensure you have the following installed on your machine:

• Python 3.6 or higher
• pip (Python package installer)

Installation

1. Install Dependencies

   pip install -r requirements.txt

Training the Model

1. Customize Intents Modify the data.json file to include your own patterns and responses.

2. Run the Training Script

   python train.py

Using tobin.py

1. Run the Tobin Script

   python tobin.py path_to_safetensors path_to_desired model.bin

   note use same folder(Fine Tuned Model)
2. Functionality The tobin.py script provides additional utilities for data preprocessing and analysis. It can be used to transform raw data into a format suitable for training and to perform exploratory data analysis.

Running the Chatbot

1. Start the Chatbot

   python chat.py

2. Interact with the Chatbot The chatbot will prompt you to enter your queries and will respond based on the trained model.

Additional Information

• Hyperparameter Tuning: You can adjust the hyperparameters in hyperparameter_tuning.py to optimize the model performance.

• Model Checkpoints: The trained model and tokenizer configurations are saved in the fine_tuned_model_checkpoint directory.

• Fine-Tuning: You can further fine-tune the model by modifying the training scripts and re-running the training process.

• Contact

• GitHub: Rayyan9477 (https://github.com/Rayyan9477)

• LinkedIn: Rayyan Ahmed (https://www.linkedin.com/in/rayyan-ahmed9477/)

• Email: rayyanahmed265@yahoo.com