🌹 Zelda Rose 🌹

A straightforward trainer for transformer-based models. In others words: a teacher of muppets.

Zelda Rose is a command line interface for pretraining transformer-based models. Its purpose is to enable an easy start for users interested in training these ubiquitous models, but unable or unwilling to engage with more comprehensive — but more complex — frameworks and the complex interactions between libraries for managing models, datasets and computations. Training a model requires no code on the user's part and produce models directly compatible with the 🤗 ecosystem, allowing quick and easy distribution and reuse. A particular care is given to lowering the cost of maintainability and future-proofing, by making the code as modular as possible and taking advantage of third-party libraries to limit ad-hoc code to the strict minimum. (Grobol, 2023)

Getting Started

Check out Get started.

Bugs and contributions

Please send them via Codeberg logo Codeberg.

Citation

If you use Zelda Rose, please cite it as :

Loïc Grobol. 2023. ‘Zelda Rose: A Tool for Hassle-Free Training of Transformer Models’. Paper presented at NLP OSS, Singapore, Indonesia. Proceedings of the 3rd Workshop for Natural Language Processing Open Source Software. https://hal.science/hal-04262806.

BibLaTeX

@inproceedings{grobol2023ZeldaRoseTool,
  title = {Zelda {{Rose}}: A Tool for Hassle-Free Training of Transformer Models},
  shorttitle = {Zelda {{Rose}}},
  booktitle = {Proceedings of the 3rd {{Workshop}} for {{Natural Language Processing Open Source Software}}},
  author = {Grobol, Loïc},
  date = {2023-12},
  url = {https://hal.science/hal-04262806},
  urldate = {2023-10-31},
  eventtitle = {{{NLP OSS}}},
}

CSL JSON

{
  "id":"grobol2023ZeldaRoseTool",
  "accessed":{"date-parts":[["2023",10,31]]},
  "author":[{"family":"Grobol","given":"Loïc"}],
  "citation-key":"grobol2023ZeldaRoseTool",
  "container-title":"Proceedings of the 3rd Workshop for Natural Language Processing Open Source Software",
  "event-place":"Singapore, Indonesia",
  "event-title":"NLP OSS",
  "issued":{"date-parts":[["2023",12]]},
  "title":"Zelda Rose: a tool for hassle-free training of transformer models",
  "type":"paper-conference",
  "URL":"https://hal.science/hal-04262806"
}