This repository contains my learning of how to make conversational AI.
This folder was the first I created. It has intents json file containing tag, patterns and answers to describe the vocabulary. It also has two python files, One for training and one for predicting.
This contains files regarding my learnnig on the usage of Keras, TensorFlow, and Tokenizers. IT contains few models such as a file to see if the statement is sarcastic, a file to do sentiment analysis, and a file to make poetry starting with a prompt based on irish folksongs.
I have created this folder to try different models of using LLMs. The first file I created was simply to use open source LLMs from huggingface transformers and use gradios library to make a user interface and render it. (Chtbt1.py) The second file I made to start learning fine-tuning of LLMs with LORA to do sentiment analysis. (finetune.py) The third file (I am still working on) is to use prompt engineering to autograde short answer type questions. I found quite lot of code was depriciated and saw a lot of documentations, however I still need to make this work.