Command line tool for improving typing speed and accuracy. The main goal is to help programmers practise programming languages.

pip install --upgrade mltype

Make sure that Docker and Docker Compose are installed.

docker-compose run --rm mltype

You will get a shell in a running container and the mlt command should be available.

See the documentation for more information.

• Using neural networks to generate text. One can use pretrained networks (see below) or train new ones from scratch.
• Alternatively, one can read text from a file or provide it manually

• Dead simple (implemented in curses)
• Basic statistics - WPM and accuracy
• Setting target speed
• Playing against past performances

• Detailed documentation: https://mltype.readthedocs.io/en/latest/index.html.
• GIF examples: https://mltype.readthedocs.io/en/latest/source/examples.html.

The entrypoint is mlt. To get information on how to use the subcommands use the --help flag (e.g. mlt file --help).

$ mlt
Usage: mlt [OPTIONS] COMMAND [ARGS]...

  Tool for improving typing speed and accuracy

Options:
  --help  Show this message and exit.

Commands:
  file    Type text from a file
  ls      List all language models
  random  Sample characters randomly from a vocabulary
  raw     Provide text manually
  replay  Compete against a past performance
  sample  Sample text from a language
  train   Train a language

See below for a list of pretrained models. They are stored on a google drive and one needs to download the entire archive.

Once you download the file, you will need to place it in ~/.mltype/languages. Note that if the folder does not exist you will have to create it. The file name can be changed to whatever you like. This name will then be used to refer to the model.

To verify that the model was downloaded succesfully, try to sample from it. Note that this might take 20 seconds the first time around.

mlt sample my_new_model

Feel free to create an issue if you want me to train a model for you. Note that you can also do it yourself easily by reading the documentation (mlt train) and getting a GPU on Google Colab (click the badge below for a ready to use notebook).

This project is very much motivated by the The Unreasonable Effectiveness of Recurrent Neural Networks by Andrej Karpathy.