IASA NLP Course

Setup Working Environment

Pre-requirements

• conda 4.12.0 (later versions may also work) - Installation
• VS Code - Ubuntu Installation
• (Optional) CUDA Version: 11.4; Driver Version: 470.129.06 - Installation

Setup environment

Poetry (Recommended)

1. Install Poetry using Poetry full guide.
   • Important: Check if it is working using poetry --version
2. Run command to keep your .venv folder right in your project: poetry config virtualenvs.in-project true
3. poetry shell
   • Important: If you have conda and 2 environments were activated: conda deactivate
4. poetry install --no-root

In order to activate environment on the next use. Important: you should be inside your project

poetry shell

Conda

If you have CUDA

conda env create -f environment_gpu.yaml

Otherwise

conda env create -f environment.yaml

In order to activate environment

conda activate iasa_nlp_env

Start Jupyter

You may use any port

jupyter lab --port 7766

Content

1. Структура та структурні елементи постановки ML задачі. Формалізація бізнес задач. Основні задачі й методи в сфері Обробки природних мов
   • Author: Sydorskyi Volodymyr
   • Recording: https://drive.google.com/drive/folders/166r0s2p8Exc7Fucs3XFkHPuQaHB3LDAQ?usp=drive_link
2. Представлення природніх мов в машинному вигляді. Класичні та нейронні алгоритми векторизації. Класичні ML підходи в NLP
   • Author: Yelisieiev Vladyslav
   • Recording: https://drive.google.com/drive/folders/1ZShCNBmnlQrvsReLdMfAy8k6oLkDK9jT?usp=drive_link
3. Основні метрики в NLP (обробка природніх мов). Побудова оцінки підходів і моделей в NLP - валідація
   • Author: Bazdyrev Anton
   • Recording: https://drive.google.com/drive/folders/1Ee-RTDhgxWCa8MPpyI29j7CIJF0t6x6s?usp=drive_link
4. Підходи з використанням архітектур RNN/GRU/LSTM
   • Author: Sydorskyi Volodymyr
   • Recording: https://drive.google.com/drive/folders/1SMGWOdwuBeN69DcGv_jbdV0zzftYAIx4?usp=drive_link
5. Підходи з використанням архітектури Transformer
   • Author: Bazdyrev Anton
   • Recording: https://drive.google.com/drive/folders/1hpdbO4ElfSt44c5MHuT-b0fhgKDSVCou?usp=drive_link
6. Генеративні задачі: машинний переклад, сумаризація тексту, умовна та безумовна текстова генерація, розгляд GPT архітектури
   • Author: Yelisieiev Vladyslav
   • Recording: https://drive.google.com/drive/folders/1uWAWtQzOaGvpkdzWWPs5YXM-T7rjLayG?usp=drive_link
7. Задача кластеризації. Задача моделювання тем
   • Author: Sydorskyi Volodymyr
   • Recording: https://drive.google.com/drive/folders/1lLx7zBQ1GnoJHP02vGaoGehaRJfG6Lwa?usp=drive_link
8. MLOps - розгортання моделей
   • Author: Bazdyrev Anton
   • Recording: https://drive.google.com/drive/folders/1TUsONxXg-RaCVuSPd5CPSsiPnqz6Nbbg?usp=drive_link

Use Kaggle or Colab for computations

Kaggle

1. Create Kaggle account
2. Create Notebook
3. Explore docs and find out how
   • Add Kaggle dataset to notebook
   • Turn on GPU

Colab

1. Create Notebook in Colab
2. Enable GPU
3. Add Kaggle dataset to Colab - https://www.geeksforgeeks.org/how-to-import-kaggle-datasets-directly-into-google-colab/

Data

• For most of lectures you will need datasets from Kaggle. Prepare in advance
  • CommonLit - Evaluate Student Summaries dataset API command: kaggle competitions download -c commonlit-evaluate-student-summaries
  • Natural Language Processing with Disaster Tweets dataset API command: kaggle competitions download -c nlp-getting-started
  • Mantis Analytics Location Detection dataset: kaggle datasets download -d vladimirsydor/mantis-analytics-location-detection
  • Dataset for Topic Modelling: https://drive.google.com/drive/folders/1jwh225T0DIEN4A1wMZ8-dVJX-2Tsovqf?usp=sharing
• We recommend to create data folder in the course root directory and put all datasets there. So you might have next structure

data/
    nlp_getting_started/
        train.csv
        test.csv
        ...
    ...
Lecture_1/
...

How to use Kaggle datasets

1. Create Kaggle account
2. Proceed with Installation & Authentication
3. Don't forget to join a competition and accept its rules on a Kaggle website.
4. Download dataset with API command

Feedback [Only For Lectors]

2023

Raw table : https://docs.google.com/spreadsheets/d/1P38uhwkMQo0cd1avywbnVJ-dwxiswHpAcIpEHrlv1PY/edit?usp=sharing

TODO

• Process recordings and upload them to YouTube
• Process 2023 Feedback

Citation

@misc{iasa_nlp_course_2023,
  author = {Sydorskyi Volodymyr, Bazdyrev Anton, Yelisieiev Vladyslav},
  title = {IASA NLP course 2023},
  year = {2023},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/VSydorskyy/iasa_nlp_course}},
}