This repository provides a comprehensive tutorial for beginners in Natural Language Processing (NLP), focusing on practical implementation with deep learning techniques. It targets new students joining the TANGENT lab, offering a structured path from foundational knowledge to hands-on projects using modern tools like PyTorch and Huggingface Transformers.

How It Works

The tutorial guides users through essential NLP concepts, including machine learning, deep learning, and specific NLP tasks. It emphasizes a "learn by doing" approach, with practical exercises covering text classification (CNN/RNN), Named Entity Recognition (LSTM-CRF), Neural Machine Translation (NMT), and Transformer models. The curriculum is designed to build a strong understanding of both theoretical underpinnings and practical application, leveraging popular libraries and frameworks.

Quick Start & Requirements

• Installation: Requires Python and environment management via Anaconda/Miniconda. PyTorch is the primary deep learning framework.
• Prerequisites: Strong Python programming skills, basic Linux experience, and a solid mathematical foundation (calculus, linear algebra, probability) are recommended. Access to GPU resources is advised for larger models, with instructions to contact the lab administrator.
• Resources: Links to Kaggle datasets, PyTorch tutorials, Huggingface documentation, and example code (e.g., ChineseNER) are provided.

Highlighted Details

• Covers foundational ML/DL concepts and their relevance to NLP.
• Practical tasks include text classification, NER, NMT, and Transformer/PLM understanding.
• Strong emphasis on using PyTorch and the Huggingface ecosystem (Transformers, Trainer).
• Guides users on literature review using Google Scholar and arXiv, and code reproduction via GitHub.

Maintenance & Community

• The repository is associated with the TANGENT lab at PKU.
• Instructions for contribution via Pull Requests and issue tracking are provided.
• Recommended reviewer is Yifan-Song793.

Licensing & Compatibility

• The repository itself does not explicitly state a license in the provided README text. Code examples may be subject to their original licenses.

Limitations & Caveats

• Some tasks may require GPU resources not available on personal computers.
• The tutorial assumes a certain level of prior knowledge in programming and mathematics.
• Users are expected to manage their own Python environments and dependencies.