Chatbot with PyTorch using Deep Learning with RNNs, LSTMs and Sequence Models with Attention Mechanism

Pytorch and Python used to crate Seq2Seq models (LSTM and GRU) for natural language processing (NLP). Furthermore, the implemented model utilized to create an AI enhanced chatbot. This project was developed at De Montfort University.

FAQ CHATBOT using pytorch LSTM Encoder-Decoder model along with beam search and greedy search

Madical ChatBot using LSTMs neural networks with Tensorflow and using fine-tuning with a pre-built model with Pytorch

Conversational chatbot using a Seq2Seq LSTM model with GloVe embeddings and PyTorch. Features include teacher forcing, temperature sampling, mixed precision training, and real-time validation.

A Seq2Seq chatbot using LSTM encoder-decoder with attention, pretrained GloVe embeddings, and beam search decoding. Built on the DailyDialog dataset as a proof-of-concept for dialog modeling with PyTorch.

Neural Language Model using LSTMs, GRUs, and Transformers (GPT, BERT) for text generation, sentiment analysis, and machine translation. Built with Python, TensorFlow/PyTorch, and Hugging Face. Supports autocomplete, chatbots, and grammar correction.

Developing a system to detect human emotions from voice inputs of elderly people, using a custom CNN+LSTM model trained with PyTorch. The system also performs text-based sentiment analysis based on the overall conversation between the elderly user and the chatbot.

Developed💬Chat with Data Chatbot⌨️ using Flask + SocketIO for natural-language CSV queries with instant text/charts. Built from scratch: PyTorch CNN (intent), TensorFlow Bi-LSTM (NER), gte-small (embeddings), FAISS (RAG), LangChain, GPT-4o-mini/Llama-3.2 (answers + chart specs), Matplotlib/Seaborn (viz). Fully local, no pretrained weight