Fiddling with GPT2 and other NLP models for interesting corpus text generation
To start run the bash script to setup the environment properly.
. ./setup_env.sh
This will initialize a git repo and make a first commit. It will also create a conda environment and kernel with the specified project name installing dependencies.
Be sure to uncomment lines:
/data/
reports/data-profiling/*
From .gitignore file if data should be ignored in versioning
To install various useful Jupyter Lab extensions, run:
. ./jupyter-extensions.sh
If the project is finished and the environment is no longer needed (it can be rebuilt running setup_env.sh again), run:
. ./teardown_env.sh
├── LICENSE
├── Makefile <- Makefile with commands like `make data` or `make train`
├── README.md <- The top-level README for developers using this project.
├── data
│ ├── external <- Data from third party sources.
│ ├── interim <- Intermediate data that has been transformed.
│ ├── processed <- The final, canonical data sets for modeling.
│ └── raw <- The original, immutable data dump.
│
├── docs <- A default Sphinx project; see sphinx-doc.org for details
│
├── models <- Trained and serialized models, model predictions, or model summaries
│
├── notebooks <- Jupyter notebooks. Naming convention is a number (for ordering),
│ the creator's initials, and a short `-` delimited description, e.g.
│ `01.0-jqp-initial-data-exploration`.
│
├── references <- Data dictionaries, manuals, and all other explanatory materials.
│
├── reports <- Generated analysis as HTML, PDF, LaTeX, etc.
│ └── figures <- Generated graphics and figures to be used in reporting
│
├── requirements.txt <- The requirements file for reproducing the analysis environment, e.g.
│ generated with `pip freeze > requirements.txt`
|
├── setup.sh <- Script to initiailize git repo, setup a conda virtual environment
│ and install dependencies.
├── teardown_env.sh <- Script to teardown the project conda virtual environment
│
├── setup.py <- For installing src as a local package
├── src <- Source code for use in this project.
│ ├── __init__.py <- Makes src a Python module
│ │
│ ├── data <- Scripts to download or generate data
│ │ └── make_dataset.py
│ │
│ ├── features <- Scripts to turn raw data into features for modeling
│ │ └── build_features.py
│ │
│ ├── models <- Scripts to train models and then use trained models to make
│ │ │ predictions
│ │ ├── predict_model.py
│ │ └── train_model.py
│ │
│ ├── visualization <- Scripts to create exploratory and results oriented visualizations
│ │ └── visualize.py
│ │
│ └── utils <- Utility code for various purposes and packages
│ ├── project_utils.py <- For project specific utilities
│ ├── gists <- Code gists with commonly used code (change to root
│ │ directory, connect to database, profile data, etc)
│ ├── io <- Code for input/output utilities
│ ├── etl <- For building reproducible ETL pipelines, including data
│ │ checks and transformers
│ ├── ml <- Machine Learning utility code (feature engineering, etc)
│ ├── pandas <- Pandas related utility code
│ │ ├── analysis
│ │ ├── cleaning
│ │ ├── engineering
│ │ ├── text
│ │ ├── datetime
│ │ ├── optimization
│ │ └── profiling
│ └── text <- Code for dealing with text. Includes distributed loading of text corpus,
│ entity statement extraction, sentiment analysis, etc.
│
├── LICENSE
│
└── tox.ini <- tox file with settings for running tox; see tox.testrun.org
Much of the boilerplate code and structure for this comes from Driven Data's Cookie Cutter Data Science