UZH_ML4NLP

This repository contains the source code of all projects developed for the ML4NLP class at the University of Zurich during fall semester 2021.
The projects cover the following topics:

1. Language Identification on Tweets using sklearn.
2. Building Word Embeddings with PyTorch or TensorFlow.
3. Language Identification on Tweets with CNNs using PyTorch and TensorFlow.
4. Paper dissection of "Improving Language Understanding by Generative Pre-Training" (GPT-1 Paper).
5. Named entity recognition by using and fine-tuning pretrained language models.
6. Topic Modeling based on Latent Dirichlet Allocation.

Useful commands

pandoc -V geometry:margin=2cm -o ex01_report.pdf README.md