NLP Samples

A collection of NLP notebooks and scripts covering classic deep learning approaches, HuggingFace transformer tasks, and LangChain LLM applications.

Deep Learning (RNN / CNN / LSTM)

File	Description
IMDB_RNN.ipynb	Sentiment analysis on IMDB reviews using RNN
IMDB_CNN.ipynb	Sentiment analysis on IMDB reviews using 1D CNN
Tweet_Emotion_LSTM.ipynb	Tweet emotion classification using LSTM (TensorFlow/Keras)
TurkceTweet_LSTM.ipynb	Turkish tweet emotion classification using LSTM
classify-emotions.ipynb	Emotion classification on Kaggle dataset

HuggingFace Transformer Tasks

Folder	Task	Model
text_classification/	Sequence classification (BERT)	BERT
token_classification/	Token classification (NER)	BERT
question_answering/	Question answering	BERT
translation/	Translation	HuggingFace
summarization/	Text summarization	BART
multiple_choice/	Multiple choice	BERT
masked_lang/	Masked language modeling	BERT
casual_lang/	Causal language modeling	GPT

LangChain Applications

File	Description
LangChain/text_generation.py	Text generation using GPT-2 via HuggingFaceHub
LangChain/question_answering.py	QA using Gemma 7B via HuggingFaceHub
LangChain/summarization.py	Summarization using BART-large-CNN
LangChain/english_to_sql.py	Natural language to SQL using T5-finetuned-wikiSQL
LangChain/vector_search.py	Semantic vector search using ChromaDB

Datasets

• smile-annotations-final.csv — Smile emotion dataset for tweet classification
• TurkishTweets.xlsx — Turkish tweets for emotion classification

Tech Stack

• Deep Learning: TensorFlow, Keras, PyTorch, d2l
• Transformers: HuggingFace Transformers (BERT, GPT, BART, Gemma)
• LLM Framework: LangChain, HuggingFaceHub
• Vector DB: ChromaDB