30 lines (25 loc) · 1.14 KB

🤖 Intelligent Chatbot Development

🔍 Project Description

I built a chatbot using transformer models trained on datasets like Cornell Movie Dialogs and Reddit chats.
It understands and replies like a human by learning context through self-attention mechanisms.
I fine-tuned it to give relevant, consistent, and tone-controlled responses using methods like beam search and sampling.

📂 Dataset Used

This project uses the Cornell Movie Dialogs Corpus, stored in the cornell_movie_dialogs folder.

🧠 The dataset includes:

movie_lines.txt
movie_conversations.txt
movie_characters_metadata.txt
movie_titles_metadata.txt
raw_script_urls.txt
chameleons.pdf
pairs_encoded.json
WORDMAP_corpus.json

These files are essential for training and testing a conversational AI system.

⚠️ Important: Git LFS Required

Some files are large and tracked using Git LFS.
If you clone this repo, make sure to install Git LFS first:

git lfs install
git clone https://github.com/Surajit00007/Intelligent_Chatbot_Development-project.git