|
--- |
|
license: cc-by-sa-4.0 |
|
language: |
|
- en |
|
size_categories: |
|
- 1M<n<10M |
|
tags: |
|
- text normalization |
|
- inverse text normalization |
|
--- |
|
|
|
This is a vocabulary of (semiotic_class, normalized, non-normalized, freq) tuples from [Google Text Normalization Dataset](https://www.kaggle.com/datasets/richardwilliamsproat/text-normalization-for-english-russian-and-polish). |
|
|
|
It was generated using [this script](https://github.com/NVIDIA/NeMo/blob/main/examples/nlp/text_normalization_as_tagging/evaluation/get_multi_reference_vocab.py). |
|
|
|
It can be used to perform fast text normalization, see example in this [script](https://github.com/bene-ges/nemo_compatible/blob/main/scripts/nlp/en_spellmapper/dataset_preparation/normalize_by_gtn_vocab.py) |
|
|
|
|
